AI background literature registry

This registry is the structured source behind the expanded background reading routes and their one-click choices in My Library. It covers mathematical and machine-learning foundations, computer vision, NLP, large language models, and the dedicated natural-image Vision & Language field.

Coverage cutoff: 2 September 2026. The catalog is a curated teaching map across the approved taxonomy and major method lineages. It is not a claim to index every paper. Route order represents prerequisites and conceptual progression—not paper quality, venue prestige, citation count, or a global model ranking.

Every detailed background chapter renders its complete route from this catalog. Shorter tables in the prose are explanatory ladders, not alternate paper lists. Routes with one main lineage progress chronologically; chapters that combine several scientific branches use an explicit prerequisite-aware sequence.

866 unique background and natural-image papers48 category filters43 detailed background routesRoutes reviewed through October 2, 2026

SearchCategory or detailed routeYearEvidence state

Showing 866 of 866 records.

2026ECCVAccepted · proceedings pending
“World Knowledge” in the Weights: Reading Concept Circuits of Vision Transformers
Yanlin Chen, Tang Li, and Xi Peng
L3
ECCV 2026 Oral verified in the official program; scientific claims use the conference-hosted manuscript #8168. Publisher proceedings/version of record was not verified on 2026-09-12.
Interpretability Vision evaluation Visual representations
interpretability / attribution · vision evaluation / robustness · visual representation learning · VisionCLT · cross-layer transcoders · concept circuits · spurious correlations · mechanistic interpretability · World Knowledge in the Weights: Reading Concept Circuits of Vision Transformers
Official artifacts: ECCV Oral program · Conference manuscript · Official implementation · arXiv manuscript
Metadata checked 2026-09-12
2026TMLRPeer reviewed
A Comparative Study of Label-free Representation Quality Metrics in Deep Learning
Daniel Richards Arputharaj, Daniel Jönsson, and Gabriel Eilertsen
L3
Published in Transactions on Machine Learning Research; the official OpenReview forum is canonical and the arXiv manuscript is linked as a supplement.
Evaluation / reproducibility Vision evaluation Visual representations
evaluation / reproducibility · vision evaluation / robustness · visual representation learning · label-free evaluation · representation quality · intrinsic dimension · model selection · benchmark study
Official artifacts: Author manuscript
Metadata checked 2026-08-30
2026arXivPreprint
A Fix for Softmax and Cross-Entropy
Jonathan T. Barron, Dan Gnanapragasam, and Michael Tschannen
L3
Inspected arXiv:2610.31642v1, submitted 22 October 2026, and the author project page. Numerical-gradient and modular-arithmetic experiments are not evidence of improved WSI or large-scale foundation-model accuracy. No venue acceptance was verified.
Loss / objectives Knowledge distillation Neural networks
loss / objective design · knowledge distillation / model transfer · neural networks · numerical stability · Softmax Collapse · Softmax Cancellation · cross-entropy · soft-target distillation · grokking · mixed precision · AdamW
Official artifacts: Inspected manuscript (v1) · Author project and JAX implementations
Metadata checked 2026-10-26
2026Cancer ResearchPeer reviewed
A Magnification Alignment Framework Enables Computation- and Communication-Efficient Computational Pathology
Chu Han, Bingchao Zhao, Tianpeng Deng, Jingqi Huang, Fangfang Liu, Jiatai Lin, Shanshan Lyu, Longfei Wang, Cheng Lu, Changhong Liang, Hannah Y. Wen, Zhenwei Shi, Xiaojing Guo, and Zaiyi Liu
L3
Peer-reviewed journal article 86(12):3060–3073.
Knowledge distillation
knowledge distillation / model transfer · MIL / WSI · WSI engineering · self-supervised learning · MAG-GLTrans · magnification alignment · registered field of view · low-magnification deployment · telepathology efficiency
Official artifacts: Author manuscript · Official code
Metadata checked 2026-08-23
2026ECCVAccepted · proceedings pending
A second-order theory of texture for depth from focus
Sreekar Sai Ranganathan and Ioannis Gkioulekas
L3
Accepted to ECCV 2026; oral presentation verified in the official program. The conference-hosted manuscript is the review authority; a separate Springer proceedings record was not verified. Author spelling follows the conference PDF. The official project still lists code and data as forthcoming.
Restoration / inverse problems 3D / spatial vision
3D / spatial vision · second-order texture · subjective speckle · depth from focus · wave optics · narrowband spectral filtering
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · Author project · arXiv record
Metadata checked 2026-09-12
2026EMNLPAccepted · proceedings pending
ACTD: Anchor-Based Cross-Tokenizer Distillation with Residual Regularization
Huiyi Zhang, Zijian Li, Xiaocheng Feng, Weitao Ma, Xiaoliang Yang, Yichong Huang, and Bing Qin
L3
The arXiv record reports acceptance to the EMNLP 2026 main conference. No final ACL Anthology proceedings record was available at the 2026-09-02 audit, so methods and results remain pinned to arXiv v1 submitted 30 August 2026.
Knowledge distillation Loss / objectives Reasoning / verification LLM systems LLM post-training Generation / MT evaluation
knowledge distillation / model transfer · loss / objective design · LLM reasoning / verification · LLM training / serving systems · LLM post-training / alignment · generation / translation evaluation · ACTD · cross-tokenizer distillation · anchor loss · residual regularization · mathematical reasoning
Official artifacts: Official implementation
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
AdaBoosting Text Prompts for Vision-Language Models
Seokhee Jin, Changhwan Sung, Sunung Mun, Hoyoung Kim, and Jungseul Ok
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #5256 checked on 2026-09-12. Publisher version of record was not independently verified.
Image–text representation Transfer / adaptation / continual Classical ML
image–text representation · classical machine learning · Text Prompt Boosting · AdaBoost · few-shot prompt ensembles
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026arXivPreprint
Adapting Without Gradients: Affine Statistics Transport and What Its Certificate Can Tell You
Salim Khazem and Ibrahim Mohamed Serouis
L3
arXiv v1 submitted 31 August 2026. The official CASTER repository was available when checked on 2 September 2026; no peer-reviewed record was verified.
Transfer / adaptation / continual Evaluation / reproducibility Vision evaluation Image / recognition
evaluation / reproducibility · vision evaluation / robustness · image recognition · CASTER · test-time adaptation · frozen models · affine transport · transportability certificate · gradient-free adaptation
Official artifacts: Official implementation
Metadata checked 2026-09-02
2026CVPRPeer reviewed
Adaptive Capacity Autoregressive Visual Tracking
Tong Lin, Yifan Bai, Shiyi Liang, Ruigang Niu, and Xing Wei
L3
Video / 3D / embodied
video / 3D / embodied vision · visual tracking · autoregressive tracking · adaptive inference
Metadata checked 2026-07-28
2026ACLPeer reviewed
Afri-MCQA: Multimodal Cultural Question Answering for African Languages
Atnafu Lambebo Tonja, Srija Anand, Emilio Villa-Cueva, Israel Abebe Azime, Jesujoba Oluwadara Alabi, Muhidin A. Mohamed, Debela Desalegn Yadeta, Negasi Haile Abadi, Abigail Oppong, Nnaemeka Casmir Obiefuna, Idris Abdulmumin, Naome A Etori, Eric Peter Wairagala, Kanda Patrick Tshinu, Imanigirimbabazi Emmanuel, Gabofetswe Malema, Alham Fikri Aji, David Ifeoluwa Adelani, and Thamar Solorio
L2–L3
Peer-reviewed ACL 2026 long paper and Best Social Impact Paper award recipient.
Multilingual / low resource Captioning / VQA V&L evaluation / safety LLM evaluation / safety
multilingual / low-resource NLP · captioning / VQA · vision–language evaluation / safety · LLM evaluation / safety · Afri-MCQA · African languages · cultural VQA · speech · multilingual evaluation
Official artifacts: Author manuscript · Official dataset
Metadata checked 2026-08-30
2026CVPRPeer reviewed
AGENTSAFE: Benchmarking the Safety of Embodied Agents on Hazardous Instructions
Zonghao Ying et al.
L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · embodied safety · hazardous instructions · agent evaluation
Metadata checked 2026-07-27
2026arXivPreprint
ALICE: Learning a General-Purpose Pathology Foundation Model from Vision, Vision-Language, and Slide-Level Experts
Jiawen Li, Tian Guan, Huijuan Shi, Xitong Ling, Mingxi Fu, Anjia Han, Chao He, and Yonghong He
L3
Knowledge distillation
foundation encoder · agglomerative distillation · multi-teacher · foundation model · knowledge distillation / model transfer
Official artifacts: Code and models
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
AnyFlow: Any-Step Video Diffusion Model with On-Policy Flow Map Distillation
Yuchao Gu, Guian Fang, Yuxin Jiang, Weijia Mao, Song Han, Han Cai, and Mike Zheng Shou
L3
ECCV 2026 Spotlight verified in the official program; scientific claims use the conference-hosted manuscript #9062. Publisher proceedings/version of record was not verified on 2026-09-12.
Knowledge distillation Generative / representation Generative vision
knowledge distillation / model transfer · generative / representation learning · generative vision · AnyFlow · flow map distillation · on-policy distillation · any-step sampling · video generation
Official artifacts: ECCV Spotlight program · Conference manuscript · Official implementation · arXiv manuscript · Author project · Model collection
Metadata checked 2026-09-12
2026CVPRPeer reviewed
ARM-Thinker: Reinforcing Multimodal Generative Reward Models with Agentic Tool Use and Visual Reasoning
Shengyuan Ding et al.
L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · agentic reward model · visual tools · evidence verification
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
Atlas is Your Perfect Context: One-Shot Customization for Generalizable Foundational Medical Image Segmentation
Ziyu Zhang, Yi Yu, Simeng Zhu, Ahmed Aly, Yunhe Gao, Ning Gu, and Yuan Xue
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #2028. Publisher proceedings/version of record was not verified on 2026-09-12.
Detection / segmentation Transfer / adaptation / continual Evaluation / reproducibility
detection / segmentation · evaluation / reproducibility · AtlasSegFM · one-shot segmentation · test-time optimization · reliability-gated fusion · cross-validation
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Bayesian Self-Attention with Intra-Token Modeling for Lightweight Denoising Transformers
Runyang He, Zuowei Shen, and Hui Ji
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #7178. Publisher proceedings/version of record was not verified on 2026-09-12. Conference program and code use 'Local Pixel Correlations'; checked conference PDF uses 'Intra-Token Modeling'. Both titles describe submission 7178; PDF title is canonical here.
Restoration / inverse problems Neural networks
neural networks · NLformer · Bayesian self-attention · intra-token correlation · Woodbury inversion · rank-2 correction · Bayesian Self-Attention with Local Pixel Correlations for Lightweight Denoising Transformers
Official artifacts: ECCV Spotlight program · Conference manuscript · Official implementation
Metadata checked 2026-09-12
2026ICLRPeer reviewed
Beyond Length: Quantifying Long-Range Information for Long-Context LLM Pretraining Data
Haoran Deng, Yingyu Lin, Zhenghao Lin, Xiao Liu, Yizhou Sun, Yian Ma, and Yeyun Gong
L3
Peer-reviewed ICLR 2026 poster; treat LongFilter as frontier evidence for long-context data curation rather than a settled recipe.
LLM pretraining
LLM pretraining / scaling · LongFilter · long-context data · data curation
Metadata checked 2026-07-28
2026CVPRPeer reviewed
BiFM: Bidirectional Flow Matching for Few-Step Image Editing and Generation
Yasong Dai, Zeeshan Hayder, David Ahmedt-Aristizabal, and Hongdong Li
L3
Generative vision
generative vision · bidirectional flow matching · few-step generation · image editing
Metadata checked 2026-07-28
2026ICLRPeer reviewed
Breaking Barriers: Do Reinforcement Post Training Gains Transfer to Unseen Domains?
Chuxuan Hu, Yuxuan Zhu, Antony Kellermann, Caleb Biddulph, Suppakit Waiwitlikhit, Jason Benn, and Daniel Kang
L3
First released as a 2025 preprint and peer-reviewed as an ICLR 2026 poster; the study is a boundary audit of RLVR transfer.
LLM post-training
LLM post-training / alignment · reinforcement post-training · RLVR · domain generalization
Metadata checked 2026-07-28
2026ECCVAccepted · proceedings pending
Broadband Full Field of View Imaging with Computational Mirrors
Vishwanath Saragadam, Niki Nezakati, Amit Roy-Chowdhury, and Vivek Boominathan
L3
Accepted to ECCV 2026; oral presentation verified in the official program. The conference-hosted manuscript is the review authority; a separate Springer proceedings record was not verified. Title follows the conference-hosted PDF; the official program/arXiv also use “Broadband Wide Field of View Imaging with Computational Mirrors”. Code/data are announced but not yet linked as released.
Restoration / inverse problems
Computational Mirrors · SeidelConv · VIS-SWIR · spatially varying PSF · focal stack · plug-and-play reconstruction · Broadband Wide Field of View Imaging with Computational Mirrors
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · Author project · arXiv record
Metadata checked 2026-09-12
2026arXivPreprint
CellDETR: A Detection-Guided Framework for Scalable Cell Representation Learning from Histopathology Images
Shikang Zhang, Guojun Li, Yicong Mao, and Chulin Sha
L3
The official implementation describes a NeurIPS 2026 submission; no acceptance or proceedings record was available when checked on 2026-08-10.
Detection / segmentation
cell / tissue modeling · foundation encoder · CellDETR · cell representation learning · Deformable DETR · histopathology · contrastive pretraining · detection / segmentation
Official artifacts: Official code
Metadata checked 2026-08-10
2026UAIPeer reviewed
Certified Interventional Fidelity: Anytime-Valid, Adaptive Evaluation of Causal Claims in Mechanistic Interpretability
Amir Asiaee
L3
Published in Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337.
Interpretability Causal inference Evaluation / reproducibility
interpretability / attribution · causal inference · evaluation / reproducibility · mechanistic interpretability · causal intervention · confidence sequence · adaptive evaluation · activation patching
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
CFM: Language-aligned Concept Foundation Model for Vision
Kai Wittenmayer, Sukrut Rao, Amin Parchami-Araghi, Bernt Schiele, and Jonas Fischer
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #4902 checked on 2026-09-12. Publisher version of record was not independently verified.
Image–text representation Interpretability Detection / segmentation
image–text representation · interpretability / attribution · detection / segmentation · concept foundation model · Matryoshka SAE · spatial concepts
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026CVPRPeer reviewed
Chain-of-Frames: Advancing Video Understanding in Multimodal LLMs via Frame-Aware Reasoning
Sara Ghazanfari et al.
L3
Generation / video / agents
multimodal generation / agents · video reasoning · frame grounding · chain of thought
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
Chain-of-Visual-Thought: Teaching VLMs to See and Think Better with Continuous Visual Tokens
Yiming Qin, Bomin Wei, Jiaxin Ge, Konstantinos Kallidromitis, Stephanie Fu, Trevor Darrell, and XuDong Wang
L3
ECCV 2026 Oral verified in the official program (event 6085); scientific claims use the conference-hosted manuscript #5952. Publisher proceedings/version of record was not verified on 2026-09-12. First preprint appeared in November 2025; stable ID preserves that origin.
Multimodal LLMs Captioning / VQA Knowledge distillation
multimodal LLM · captioning / VQA · knowledge distillation / model transfer · continuous visual reasoning · perceptual expert distillation · latent tokens
Official artifacts: ECCV program · Conference manuscript · Author project · Code
Metadata checked 2026-09-12
2026ACLPeer reviewed
Characterizing the Expressivity of Local Attention in Transformers
Jiaoda Li and Ryan Cotterell
L2–L3
Peer-reviewed ACL 2026 long paper and Best Paper award recipient.
Transformers / pretraining LLM pretraining
NLP transformers / pretraining · LLM pretraining / scaling · local attention · global attention · formal languages · fixed precision · Transformer expressivity
Official artifacts: Author manuscript
Metadata checked 2026-08-30
2026ACLPeer reviewed
CodeFlowBench: A Multi-turn, Iterative Benchmark for Complex Code Generation
Sizhe Wang, Zhengren Wang, Dongsheng Ma, Yongan Yu, Rui Ling, Zhiyu Li, Feiyu Xiong, and Wentao Zhang
L2–L3
Code / software engineering
code generation / software engineering · CodeFlowBench · multi-turn code generation · dependency-aware generation · code reuse · repository code
Official artifacts: Official implementation and benchmark data
Metadata checked 2026-08-20
2026arXivPreprint
CODI: Improving Self-Supervised Vision Transformers with Cross Distillation
Siran Dai, Qianqian Xu, Peisong Wen, Yang Liu, Xiaochun Cao, and Qingming Huang
L3
arXiv:2610.30639v1, submitted 22 October 2026, inspected 26 October 2026. The manuscript has a NeurIPS 2026 footer, but no official acceptance record was independently located in this audit; retained as a preprint. The author repository contains only a Coming soon README, not released implementation code.
Knowledge distillation Loss / objectives Generative / representation Visual representations Detection / segmentation
knowledge distillation / model transfer · loss / objective design · generative / representation learning · visual representation learning · detection / segmentation · CODI · cross distillation · self-supervised learning · DINO · iBOT · class-to-patch target mixing · dense representations · EMA teacher
Official artifacts: Inspected manuscript (v1) · Author repository — implementation pending at inspection
Metadata checked 2026-10-26
2026arXivPreprint
Computationally Efficient Pathology Segmentation using Knowledge Distillation from Foundation Models
Jiaqi Lv, Yijie Zhu, Saki Okada, Maxwell Stanley Renna, Robert Goldin, Stefan Antonowicz, and Shan E Ahmed Raza
L3
arXiv v1 (2026-09-03), submitted to Medical Image Analysis according to the authors, not accepted. Verified methods and selected results on 2026-10-02. TIAToolbox is the stated release destination; availability of these exact weights was not independently confirmed. Patch-throughput measurements are not end-to-end WSI latency.
Knowledge distillation Detection / segmentation
cell / tissue modeling · foundation encoder · WSI engineering · knowledge distillation / model transfer · detection / segmentation · Virchow2 · segmentation knowledge distillation · pixel and feature transfer · SegFormer · EUNet · efficient WSI processing
Official artifacts: Versioned manuscript · TIAToolbox — stated release destination
Metadata checked 2026-10-02
2026MIDLPeer reviewed
ConStruct: Structural Distillation of Foundation Models for Prototype-Based Weakly Supervised Histopathology Segmentation
Khang Le, Ha Thach, Anh M. Vu, Trang T. K. Vo, Han H. Huynh, David Yang, Minh H. N. Le, Thanh-Huy Nguyen, Akash Awasthi, Chandra Mohan, Zhu Han, and Hien Van Nguyen
L3
Peer-reviewed MIDL 2026 proceedings, PMLR 315:2893–2905; official record verified 2026-10-02.
Knowledge distillation
cell / tissue modeling · pathology VLM · knowledge distillation / model transfer · ConStruct · structural knowledge distillation · CONCH · SegFormer · weakly supervised segmentation · text-guided prototypes
Official artifacts: Official proceedings PDF · Official implementation
Metadata checked 2026-10-02
2026ACLPeer reviewed
CoSToM: Causal-oriented Steering for Intrinsic Theory-of-Mind Alignment in Large Language Models
Mengfan Li, Xuanhua Shi, and Yang Deng
L2–L3
Peer-reviewed ACL 2026 long paper and Best Theme Paper award recipient.
LLM interpretability LLM post-training Reasoning / verification
LLM interpretability / editing / unlearning · LLM post-training / alignment · LLM reasoning / verification · CoSToM · theory of mind · causal tracing · activation steering · social reasoning
Official artifacts: Author manuscript · Official code
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
Cross-Resolution Distribution Matching for Diffusion Distillation
Feiyang Chen, Hongpeng Pan, Haonan Xu, Xinyu Duan, Zhefeng Wang, and Yang Yang
L3
ECCV 2026 poster verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. RMD retains high-resolution refinement at inference; it is not low-resolution-only recognition. Reported speedups depend on the single-A100 baseline, CFG and NFE. The conference PDF author order differs from the arXiv record and is preserved here.
Knowledge distillation Generative / representation Generative vision Generation / video / agents
knowledge distillation / model transfer · generative / representation learning · generative vision · multimodal generation / agents · RMD cross-resolution · distribution matching distillation · logSNR alignment · image and video generation · multi-resolution inference
Official artifacts: Official ECCV poster record
Metadata checked 2026-09-12
2026Engineering Applications of Artificial IntelligencePeer reviewed
Cross-Stain Knowledge Distillation for Low-Cost Lung Cancer Programmed Death Ligand-1 Assessment With Multi-Granularity Multiple Instance Learning
Yi Shi, Chong Ge, Fang Zhao, Anli Zhang, Ao Li, Haibo Wu, and Minghui Wang
L3
Peer-reviewed journal article, volume 167, article 113762. IHC is privileged teacher information during training; deployment uses the H&E student, but its macro- and micro-granularity branches remain. No official code or model release was verified on 2026-09-02.
Knowledge distillation
knowledge distillation / model transfer · MIL / WSI · spatial / molecular · WSI engineering · evaluation · CroSMuG · cross-stain distillation · PD-L1 assessment · privileged IHC teacher · multi-granularity MIL
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Cross-View Yaw Estimation in Location Uncertainty with Line-Aligning Yaw Scoring
Taeho Kang, Nairan Zhang, Yelin Kim, Yujiao Shi, and Youngki Lee
L3
Accepted to ECCV 2026; oral presentation verified in the official program. The conference-hosted manuscript is the review authority; a separate Springer proceedings record was not verified. Exact KITTI measurements follow Table 2, which differs from the surrounding prose; that discrepancy is retained in the review.
3D / spatial vision Video / 3D / embodied
3D / spatial vision · video / 3D / embodied vision · LAYS · cross-view localization · yaw estimation · radial invariance · line consensus voting
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · Author project · arXiv record
Metadata checked 2026-09-12
2026arXivPreprint
CSCWD: Cross-Scale Channel-wise Knowledge Distillation for Lightweight Tiny Object Detection on Edge Devices
Amir Zamani and Zeinab Ghasemi-Naraghi
L3
arXiv v1 checked 2 October 2026; no official code release or acceptance verified. Cross-scale means teacher P2 to student P3 feature-map alignment at a shared input resolution, not independent high/low magnification acquisitions. MGD is stronger on reported mAP; external stricter mAP does not improve.
Knowledge distillation Detection / segmentation
knowledge distillation / model transfer · detection / segmentation · channel-wise-kd · knowledge-distillation
Metadata checked 2026-10-02
2026arXivPreprint
Decoupled I/O-Dominant Pipelines for Large-Scale Whole-Slide Image Embedding Extraction
Mayanka Chandrashekar, Xi Zhang, Ethan Seefried, Tirthankar Ghosal, John Gounley, and Heidi Hanson
L3
arXiv v1 systems preprint submitted 27 August 2026; no peer-reviewed version, official implementation, or reusable deployment package was verified by the review cutoff.
Evaluation / reproducibility
WSI engineering · evaluation / reproducibility · WSI embedding extraction · I/O bottleneck · decoupled pipeline · distributed inference · vector database · systems scaling
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
Delving into Latent Spectral Biasing of Video VAEs for Superior Diffusability
Shizhan Liu, Xinran Deng, Zhuoyi Yang, Jiayan Teng, Xiaotao Gu, and Jie Tang
L3
ECCV 2026 Oral verified in the official program; scientific claims use the conference-hosted manuscript #7577. Publisher proceedings/version of record was not verified on 2026-09-12. The initial arXiv manuscript appeared in December 2025; year follows the 2026 conference.
Generative / representation Loss / objectives Generative vision
generative / representation learning · loss / objective design · generative vision · SSVAE · video tokenizer · spectral regularization · latent masked reconstruction · local correlation
Official artifacts: ECCV Oral program · Conference manuscript · Official implementation · arXiv manuscript · Model checkpoints
Metadata checked 2026-09-12
2026ICLRPeer reviewed
Detecting Data Contamination in LLMs via In-Context Learning
Michał Zawalski, Meriem Boubdir, Klaudia Bałazy, Besmira Nushi, and Pablo Ribalta
L3
Peer-reviewed ICLR 2026 poster; CoDeC is a contamination-detection instrument whose assumptions still require task-specific validation.
LLM evaluation / safety
LLM evaluation / safety · CoDeC · data contamination · in-context learning
Metadata checked 2026-07-28
2026TMLRPeer reviewed
DINOv3
Meta AI et al.
Research
Visual representations
self-supervised learning · DINOv3 · scaling · visual representation learning
Metadata checked 2026-07-27
2026arXivPreprint
Distill the Visual Evidence, Not Just the Answer: Cross-World On-Policy Distillation for Vision-Language Models
Yuanhao Sun, Huawei Ji, Jiaxin Ding, Luoyi Fu, and Xinbing Wang
L3
Preprint checked 2026-10-02. The paper calls the method CW-OPD; its linked code repository uses the name CWAD. No conference acceptance inferred.
Knowledge distillation Loss / objectives Multimodal LLMs Captioning / VQA V&L evaluation / safety
knowledge distillation / model transfer · loss / objective design · multimodal LLM · captioning / VQA · vision–language evaluation / safety · CW-OPD · counterfactual visual evidence · on-policy distillation · cross-world transitions · CWBench
Official artifacts: Versioned full text · Author implementation (repository name CWAD)
Metadata checked 2026-10-02
2026ICLRPeer reviewed
Distilled Pretraining: A modern lens of Data, In-Context Learning and Test-Time Scaling
Sachin Goyal, David Lopez-Paz, and Kartik Ahuja
L3
LLM pretraining Reasoning / verification Knowledge distillation
LLM pretraining / scaling · LLM reasoning / verification · knowledge distillation / model transfer · distilled pretraining · soft-label distillation · IsoData · token routing · in-context learning · test-time scaling
Official artifacts: Proceedings PDF · arXiv
Metadata checked 2026-10-12
2026CVPRPeer reviewed
Distilling Balanced Knowledge from a Biased Teacher
Seonghak Kim
L2–L3
Knowledge distillation Loss / objectives Image / recognition
knowledge distillation / model transfer · loss / objective design · image recognition · long-tailed recognition · biased teacher · group balancing · logit distillation · class imbalance
Metadata checked 2026-08-10
2026UAIPeer reviewed
Distilling Safe LLM Systems via Soft Prompts for On Device Settings
Motasem Alfarra, Cristina Pinneri, Dana Kianfar, Mohammed Almousa, and Christos Louizos
L3
Published in Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337.
Knowledge distillation Transfer / adaptation / continual LLM evaluation / safety
knowledge distillation / model transfer · LLM evaluation / safety · safety distillation · soft prompts · on-device LLM · guard model · parameter-efficient adaptation
Metadata checked 2026-08-30
2026arXivPreprint
Divergence-Routed On-Policy Distillation for Language Models
Tiantian Peng, Guannan Lv, Hongyang Li, Qiuhe Hong, Hang Cheng, Hongjian Dou, Xiangyu Wu, and Yuyang Liu
L3
Review pins arXiv v1 (14 October 2026). The manuscript says under review at ICLR 2027; this is not acceptance.
Knowledge distillation LLM post-training Reasoning / verification
knowledge distillation / model transfer · LLM post-training / alignment · LLM reasoning / verification · DrOPD · on-policy distillation · Jensen–Shannon divergence · forward KL · reverse KL · mathematical reasoning
Official artifacts: Version 1 paper and appendices
Metadata checked 2026-10-26
2026arXivPreprint
Do As I Do, Not As I Say: On-Policy Distillation Inherits the Teacher in Vision-Language Models
Frank F. Yang, Kefei Yan, Yiren Wang, and Yiyang Wang
L3
Version 1 empirical study. The current author README describes an anonymized code-only supplement and explicitly excludes evaluation records, logs, adapters, tests and figure scripts despite the manuscript release statement.
Knowledge distillation Multimodal LLMs V&L evaluation / safety
knowledge distillation / model transfer · multimodal LLM · vision–language evaluation / safety · on-policy distillation · teacher reliability · completion failure · ChartQA · Geometry3K · filtered SFT
Official artifacts: Version 1 paper and appendices · Author code and current availability scope
Metadata checked 2026-10-26
2026ECCVAccepted · proceedings pending
Do Not Leave a Gap: Hallucination-Free Object Concealment in Vision-Language Models
Amira Guesmi and Muhammad Shafique
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #12081 checked on 2026-09-12. Publisher version of record was not independently verified.
V&L evaluation / safety Vision evaluation Grounding / open vocabulary
vision–language evaluation / safety · vision evaluation / robustness · grounding / open vocabulary · Background-Consistent Re-encoding · object concealment · grounded hallucination · adversarial robustness
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026arXivPreprint
Do We Really Need KL Divergence for On-Policy Distillation of Large Language Models?
Wenze Lin, Jiyuan Long, Jiale Zhao, Shenzhi Wang, Xitai Jiang, Ce Luo, Rui Lan, Qianli Ma, Fukang Wen, Hui Wu, Liyuan Chen, Shuoling Liu, Jiangpeng Yan, and Gao Huang
L3
Preprint checked 2026-10-02. Empirical results concern autoregressive math and code distillation; the title is not a general theorem eliminating KL from all distillation.
Knowledge distillation Loss / objectives Generation / MT evaluation LLM post-training Reasoning / verification Code / software engineering
knowledge distillation / model transfer · loss / objective design · generation / translation evaluation · LLM post-training / alignment · LLM reasoning / verification · code generation / software engineering · BinaryOPD · C-MOPD · on-policy distillation · directional reward · multi-teacher consensus
Official artifacts: Versioned full text · Author implementation
Metadata checked 2026-10-02
2026ICLRPeer reviewed
DTO-KD: Dynamic Trade-off Optimization for Effective Knowledge Distillation
Zeeshan Hayder, Ali Cheraghian, Lars Petersson, Mehrtash Harandi, Richard Hartley
L3
Official ICLR 2026 oral and proceedings manuscript checked. Adaptive task/distillation optimization, not cross-magnification or residual-error KD. Separate the exact two-gradient construction from the amortized implementation; the printed simplex handling and lower-bound argument require qualification, documented in the source-checked review.
Knowledge distillation Loss / objectives Image / recognition Detection / segmentation
knowledge distillation / model transfer · loss / objective design · image recognition · detection / segmentation · DTO-KD · ICLR 2026 oral · gradient conflict · gradient dominance · multi-objective optimization · adaptive loss weighting · amortized optimization
Official artifacts: Official proceedings record · OpenReview · Theoretical supplement
Metadata checked 2026-09-19
2026ECCVAccepted · proceedings pending
EmbedCopilot-Bench: Evaluating Vision-Language Models for Hardware-Aware Embedded System Development
Dongsheng Yuan, Yimo Deng, and Huangxun Chen
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #9127 checked on 2026-09-12. Publisher version of record was not independently verified. The PDF uses EmbedCopilot-Bench; the program shortens this to EmbedCopilot.
V&L evaluation / safety Code / software engineering Generation / video / agents
vision–language evaluation / safety · code generation / software engineering · multimodal generation / agents · embedded development · execution success · hardware grounding · EmbedCopilot: Evaluating Vision-Language Models for Hardware-Aware Embedded System Development
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Event-based Sparse-view Background-Oriented Schlieren Tomography
Xinyu Zhou, Shihao Hu, Peiqi Duan, Chao Xu, and Boxin Shi
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #377. Publisher proceedings/version of record was not verified on 2026-09-12.
Restoration / inverse problems 3D / spatial vision
3D / spatial vision · EventBOS · neural fluid field · physics-informed inverse rendering
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026ACLPeer reviewed
Evolutionary Guided Decoding: Iterative Value Refinement for LLMs
Zhenhua Liu, Lijun Li, Ruizhe Chen, Yuxian Jiang, Tong Zhu, Zhaochen Su, Wenliang Chen, and Jing Shao
L2–L3
Peer-reviewed ACL 2026 long paper and Outstanding Paper award recipient.
Generation / MT evaluation LLM post-training Reasoning / verification
generation / translation evaluation · LLM post-training / alignment · LLM reasoning / verification · evolutionary guided decoding · value refinement · on-policy evaluation · reranking · test-time generation
Official artifacts: Earlier author manuscript
Metadata checked 2026-08-30
2026IEEE Transactions on Neural Networks and Learning SystemsPeer reviewed
Expandable Residual Approximation for Knowledge Distillation
Zhaoyi Yan, Binghui Chen, Yunfan Liu, and Qixiang Ye
L3
Published in IEEE TNNLS 37(1), January 2026, pages 191–203; first available online in 2025.
Knowledge distillation Image / recognition Detection / segmentation Visual representations
knowledge distillation / model transfer · image recognition · detection / segmentation · visual representation learning · ERA · expandable residual approximation · MBRNet · teacher weight integration · feature distillation
Official artifacts: Author manuscript · Official implementation and models
Metadata checked 2026-09-02
2026ECCV eXCV WorkshopAccepted · proceedings pending
EXPOSE: Explainable and Domain-Robust Embeddings from Pathology Vision Foundation Models using Sparse Autoencoders
Anja Witte, Maximilian Lennartz, Jan Baumbach, Guido Sauter, Stefan Bonn, Patrick Fuhlert, and Marina Zimmermann
L3
Independently verified on the official eXCV 2026 Proceedings Track list, not ECCV main conference. Scientific review uses arXiv:2608.28191v1; a final typeset proceedings version was not verified. Evaluation is prostate tissue-microarray spots, not a whole-slide benchmark, and includes a thin-section failure case.
Interpretability
robustness / domain shift / privacy · evaluation · interpretability / attribution · SAE · EXPOSE · sparse autoencoders · domain feature masking · H0-mini · TMA · domain robustness
Official artifacts: Official workshop accepted-paper list · Author implementation · Versioned full manuscript and appendix
Metadata checked 2026-10-09
2026CVPRPeer reviewed
FALCON: False-Negative Aware Learning of Contrastive Negatives in Vision-Language Alignment
Myunsoo Kim, Seongwoong Shim, and Byung-Jun Lee
L3
Image–text representation Loss / objectives
image–text representation · false negatives · negative mining · contrastive learning · loss / objective design
Metadata checked 2026-07-27
2026arXivPreprint
Fast Weight Attention for Continual Learning
Yifan Zhang, Steve Ta, Jasper Zhang, Jichen Feng, Shuzhen Li, Yongxin Zhang, Yifeng Liu, Huizhuo Yuan, Mengdi Wang, Quanquan Gu, and Andrew Chi-Chih Yao
L3
arXiv v1 submitted 27 August 2026. The official implementation was available when checked on 2 September 2026; no peer-reviewed record was verified.
Text / sequence models Transformers / pretraining LLM pretraining Transfer / adaptation / continual
text / sequence modeling · NLP transformers / pretraining · LLM pretraining / scaling · fast weights · continual learning · read-after-write · NLMS · linear attention · length extrapolation
Official artifacts: Official implementation
Metadata checked 2026-09-02
2026ICLRPeer reviewed
FlashVID: Efficient Video Large Language Models via Training-free Tree-based Spatiotemporal Token Merging
Ziyang Fan, Keyu Chen, Ruilong Xing, Yulin Li, Li Jiang, Zhuotao Tian
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Inference-time token selection/merging, not residual-error KD. Full configuration includes segmentation, budget adjustment and hybrid LLM pruning. Report encoding and compression overhead as well as LLM savings; natural-video evidence is not cross-magnification pathology validation.
LLM systems Multimodal LLMs Video / 3D / embodied
LLM training / serving systems · multimodal LLM · video / 3D / embodied vision · FlashVID · video token compression · spatiotemporal merging · training-free · diverse token selection · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview · Official implementation
Metadata checked 2026-09-19
2026ICLRPeer reviewed
FlashWorld: High-quality 3D Scene Generation within Seconds
Xinyang Li, Tengfei Wang, Zixiao Gu, Shengchuan Zhang, Chunchao Guo, Liujuan Cao
L3
Official ICLR 2026 oral and full proceedings manuscript checked. A frozen multi-view teacher guides a few-step rendered-Gaussian student using learned fake scores, consistency and OOD co-training. Shared-scene rendering is not guaranteed correct geometry. Preserve different benchmark timing protocols, proprietary-data and training-code access limitations. No WSI or cross-magnification evidence.
Knowledge distillation Generative vision 3D / spatial vision
knowledge distillation / model transfer · generative vision · 3D / spatial vision · FlashWorld · cross-mode distillation · DMD2 · 3D Gaussian generation · few-step generation · distribution matching · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview · Official implementation
Metadata checked 2026-09-19
2026ECCVAccepted · proceedings pending
Fourier Compressor: Frequency-Domain Visual Token Compression for Vision-Language Models
Huanyu Wang, Jushi Kai, Haoli Bai, Lu Hou, Bo Jiang, Ziwei He, Zhouhan Lin
L3
ECCV 2026 spotlight verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. Parameter-free compression is NOT training-free: Section 5.1 and Table 2 require adaptation. The 31.2% number concerns reduced time to first token, not a measured 31.2% increase in autoregressive decoding throughput.
Multimodal LLMs Image–text representation Generation / video / agents LLM systems
multimodal LLM · image–text representation · multimodal generation / agents · LLM training / serving systems · Fourier Compressor · DCT · frequency-domain token compression · image and video · adaptation training · ECCV 2026 spotlight
Official artifacts: Official ECCV spotlight record · Official code
Metadata checked 2026-09-12
2026ACLPeer reviewed
From Local to Global: Revisiting Structured Pruning Paradigms for Large Language Models
Ziyan Wang, Enmao Diao, Qi Le, Pu Wang, Minwoo Lee, Shu-ping Yeh, Evgeny Stupachenko, Hao Feng, and Li Yang
L2–L3
Peer-reviewed ACL 2026 long paper and Outstanding Paper award recipient.
LLM systems LLM evaluation / safety
LLM training / serving systems · LLM evaluation / safety · GISP · structured pruning · global ranking · attention heads · MLP channels
Official artifacts: Author manuscript
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
GAP-Track: Bridging the Resolution Gap for Cross-Resolution RGBT Tracking
Shiyu Zhang, Tianyang Xu, Zhangyong Tang, He Wang, Xiao-Jun Wu, and Josef Kittler
L3
ECCV 2026 poster verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. Table 1 governs the recreated LasHeR values; the quarter-resolution CKD SR differs between Table 1 (35.3) and Table 4 (35.5). The PGS angle/radial equations need origin and quadrant safeguards not specified in the displayed formulas. The repository contains implementation directories despite a remaining release-soon notice; code was inspected, not executed.
Knowledge distillation Loss / objectives Video / 3D / embodied
knowledge distillation / model transfer · loss / objective design · video / 3D / embodied vision · GAP-Track · cross-resolution distillation · RGBT tracking · residual feature reconstruction · teacher assistant · low-resolution-only inference
Official artifacts: Official ECCV poster record · Official code
Metadata checked 2026-09-12
2026UAIPeer reviewed
Gaussian Process Limit Reveals Structural Benefits of Graph Transformers
Nil Ayday, Lingchu Yang, and Debarghya Ghoshdastidar
L3
Published in Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337.
Graph / geometric ML Neural networks
graph / geometric learning · neural networks · graph transformer · Gaussian process limit · infinite width · community structure · oversmoothing
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
GaussianGPT: Towards Autoregressive 3D Gaussian Scene Generation
Nicolas von Lützow, Barbara Rössle, Katharina Schmid, and Matthias Nießner
L3
Accepted to ECCV 2026; oral presentation verified in the official program. This record and review use the conference-hosted manuscript checked 2026-09-12; a separate Springer proceedings record was not verified.
3D / spatial vision Generative vision
3D / spatial vision · generative vision · GaussianGPT · 3D Gaussian generation · autoregressive modeling · lookup-free quantization · scene completion · 3D RoPE
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · arXiv record · Official code · Official project
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Geometric Context Transformer for Streaming 3D Reconstruction
Lin-Zhuo Chen, Jian Gao, Shangzhan Zhang, Yihang Chen, Nan Xue, Jianyuan Wang, Christian Rupprecht, Xun Cao, Xing Zhu, Yujun Shen, Yao Yao, and Yinghao Xu
L3
Accepted to ECCV 2026; oral presentation verified in the official program. This record and review use the conference-hosted manuscript checked 2026-09-12; a separate Springer proceedings record was not verified. The arXiv/released system uses the LingBot-Map name and a different author list; authors and quantitative claims here follow the official ECCV manuscript.
3D / spatial vision Video / 3D / embodied
3D / spatial vision · video / 3D / embodied vision · GCT · LingBot-Map · streaming reconstruction · geometric context attention · trajectory memory · KV cache
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · arXiv record · Official code (LingBot-Map)
Metadata checked 2026-09-12
2026ICMLPeer reviewed
Geometry-Aware Image Flow Matching
Junho Lee, Kwanseok Kim, and Joonseok Lee
L3
Generative / representation Graph / geometric ML Generative vision
generative / representation learning · graph / geometric learning · generative vision · SOT-CFM · Spherical Flow Matching · angular transport · hyperspherical projection · geodesic interpolation · image generation
Official artifacts: Final proceedings PDF · Author manuscript
Metadata checked 2026-10-12
2026arXivPreprint
GigaPath-Flash and GigaTIME-Flash: Efficient Pathology Foundation Models for Whole-Slide and Tumor Microenvironment Analysis
Naoto Usuyama et al.
L3
Version 2 dated 2026-07-22; author-reported performance remains provisional.
Knowledge distillation
self-supervised learning · foundation encoder · MIL / WSI · spatial / molecular · efficient foundation model · distillation · tumor microenvironment · knowledge distillation / model transfer
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
GRADE: Benchmarking Discipline-Informed Reasoning in Image Editing
Mingxin Liu, Ziqian Fan, Zhaokai Wang, Leyao Gu, Zirun Zhu, Yiguo He, Yuchen Yang, Changyao Tian, Xiangyu Zhao, Ning Liao, Shaofeng Zhang, Qibing Ren, Zhihang Zhong, Xuanhe Zhou, Junchi Yan, and Xue Yang
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #2706 checked on 2026-09-12. Publisher version of record was not independently verified.
Generative vision Generation / video / agents V&L evaluation / safety Document intelligence
generative vision · multimodal generation / agents · vision–language evaluation / safety · document intelligence · GRADE · discipline-informed image editing · scientific diagrams · reasoning evaluation
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026CVPRPeer reviewed
Grounding Everything in Tokens for Multimodal Large Language Models
Xiangxuan Ren et al.
L3
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · multimodal LLM · tokenized grounding · spatial output
Metadata checked 2026-07-27
2026CVPRPeer reviewed
GroundingME: Exposing the Visual Grounding Gap in MLLMs through Multi-Dimensional Evaluation
Rang Li et al.
L3
Grounding / open vocabulary V&L evaluation / safety
grounding / open vocabulary · vision–language evaluation / safety · grounding benchmark · rejection · spatial relations
Metadata checked 2026-07-27
2026MICCAIPeer reviewed
Harnessing Adversarial Distillation to Customise Debiased, Disease-Specific Pathology Foundation Models for Breast Cancer
Zhiwei Chen, Yang Hu, Yuxiang Xiao, Yakun Ju, Tianyang Zhang, Yingxue Xu, Wei Li, Hao Chen, Jens Rittscher, and Kaixiang Yang
L3
Official MICCAI 2026 proceedings record and accepted-version open-access PDF verified 2026-10-02; Springer DOI and pagination remain pending.
Knowledge distillation
foundation encoder · self-supervised learning · evaluation · SmartStu · adversarial distillation · disease-specific foundation model · debiasing · knowledge distillation / model transfer
Official artifacts: Official code · Official MICCAI 2026 proceedings · Official accepted-version PDF
Metadata checked 2026-10-02
2026NeurIPSAccepted · proceedings pending
HEED: Density-Weighted Residual Alignment for Hybrid Vision-Language Model Distillation
Yihao Liang and Niraj K. Jha
L3
Official NeurIPS 2026 program snapshot lists Accept (poster), event 150667 and forum gNV0DzH5iz, checked 8 October 2026. Review uses arXiv v1; final proceedings and a public author implementation were not verified. Section 4.1 and Appendix C.3 disagree about Stage 3 trainable parameters. Inference speed belongs to the hybrid architecture, not HEED weighting; no WSI magnification transfer was tested.
Knowledge distillation Loss / objectives Multimodal LLMs V&L evaluation / safety
knowledge distillation / model transfer · loss / objective design · multimodal LLM · vision–language evaluation / safety · HEED · density-weighted residual alignment · hybrid VLM · knowledge distillation · fine-grained visual perception
Official artifacts: Versioned full text · Official acceptance identity · Official NeurIPS program
Metadata checked 2026-10-08
2026NeurIPSAccepted · proceedings pending
Hide to See: Reasoning-prefix Masking for Visual-anchored Thinking in VLM Distillation
Seonghoon Yu, Dongjun Nam, Byung-Kwan Lee, and Jeany Son
L3
NeurIPS 2026 acceptance is author-reported by NVIDIA Research and the official implementation, checked 8 October 2026. No exact title or author match was found in the dated official oral/poster snapshot; presentation tier and final proceedings are unverified. Review uses arXiv v4. Training-time textual-prefix masking is not WSI magnification transfer.
Knowledge distillation Multimodal LLMs Captioning / VQA
knowledge distillation / model transfer · multimodal LLM · captioning / VQA · Masking-KD · reasoning-prefix masking · visual grounding · knowledge distillation · reverse KL
Official artifacts: Versioned full text · Official implementation · Author project · Author acceptance record
Metadata checked 2026-10-08
2026WACVPeer reviewed
HistoMILKD: A Multiple Instance Learning based Multi-Teacher Knowledge Distillation Framework for Whole Slide Image Classification
Mayur Mallya, Ali Khajegili Mirabadi, Hossein Farahani, and Ali Bashashati
L2–L3
Knowledge distillation
MIL / WSI · knowledge distillation · multi-teacher · foundation encoders · knowledge distillation / model transfer
Official artifacts: Code
Metadata checked 2026-07-27
2026ACLPeer reviewed
HSCodeComp: A Realistic and Expert-level Agent Benchmark for Hierarchical Rule Application
Tian Lan, Yiqian Yang, Qianghuai Jia, Li Zhu, Hui Jiang, Hang Zhu, Weihua Luo, and Longyue Wang
L2–L3
Peer-reviewed ACL 2026 long paper and Best Resource Paper award recipient; later label-audit concerns should accompany benchmark use.
RAG / tools / agents Reasoning / verification LLM evaluation / safety Generation / video / agents V&L evaluation / safety
RAG / tools / agents · LLM reasoning / verification · LLM evaluation / safety · multimodal generation / agents · vision–language evaluation / safety · HSCodeComp · agent benchmark · hierarchical rules · tariff classification · reasoning drift
Official artifacts: Official benchmark repository · Official dataset · Independent primary label-audit manuscript
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
Ice Cloud Geometry Retrieval with Calibrated Uncertainty from Passive Satellite Imagery
Ayush Prasad
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #12985. Publisher proceedings/version of record was not verified on 2026-09-12. Selected aggregate values follow Table 1. Table 2 reports a different G-Q aggregate R2 (0.725 vs 0.731); keep table-specific provenance. CQR guarantees marginal coverage only under exchangeability, not arbitrary subgroups or global domain shift.
Restoration / inverse problems Vision evaluation Evaluation / reproducibility
vision evaluation / robustness · evaluation / reproducibility · Sparse2Cloud · ice cloud geometry · sparse supervision · conformalized quantile regression
Official artifacts: ECCV Spotlight program · Conference manuscript · Official implementation · Author project (paper/program link)
Metadata checked 2026-09-12
2026ACLPeer reviewed
ImplicitMemBench: Measuring Unconscious Behavioral Adaptation in Large Language Models
Chonghan Qin, Xiachong Feng, Weitao Ma, Xiaocheng Feng, and Lingpeng Kong
L2–L3
Peer-reviewed ACL 2026 long paper and Best Resource Paper award recipient.
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · ImplicitMemBench · procedural memory · priming · conditioning · behavioral adaptation
Official artifacts: Author manuscript · Official benchmark code
Metadata checked 2026-08-30
2026CVPRPeer reviewed
Improving Vision-language Models with Perception-centric Process Reward Models
Yingqian Min et al.
L3
Multimodal LLMs Grounding / open vocabulary V&L evaluation / safety
multimodal LLM · grounding / open vocabulary · vision–language evaluation / safety · process reward · visual claims · reinforcement learning
Metadata checked 2026-07-27
2026CVPRPeer reviewed
In Pursuit of Pixel Supervision for Visual Pre-training
Yuwen Yang et al.
L2–L3
Visual representations Detection / segmentation
visual representation learning · detection / segmentation · Pixio · pixel supervision · dense visual pretraining
Official artifacts: Official code · Official model
Metadata checked 2026-08-08
2026ICLRPeer reviewed
InfoNCE Induces Gaussian Distribution
Roy Betser, Eyal Gofer, Meir Yossef Levi, Guy Gilboa
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Conditional theory and representation diagnostics, not proof that every InfoNCE encoder is jointly Gaussian. Separate fixed-dimensional projection limits from a regularized population surrogate and coordinate-wise non-rejection from joint normality.
Generative / representation Loss / objectives Visual representations
generative / representation learning · loss / objective design · visual representation learning · InfoNCE · contrastive learning theory · Gaussian representation · alignment uniformity · spherical concentration · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview
Metadata checked 2026-09-19
2026CVPRPeer reviewed
InternVideo-Next: Towards World-Understanding Video Models
Yi Wang et al.
L3
Video / 3D / embodied
video / 3D / embodied vision · InternVideo-Next · video world model · masked learning
Metadata checked 2026-07-27
2026CVPRPeer reviewed
Interpretable and Steerable Concept Bottleneck Sparse Autoencoders
Akshay Kulkarni, Tsui-Wei Weng, Vivek Narayanaswamy, Shusen Liu, Wesam A. Sakla, and Kowshik Thopalli
L3
Image–text representation V&L evaluation / safety Multimodal LLMs Interpretability
image–text representation · vision–language evaluation / safety · multimodal LLM · interpretability / attribution · SAE · CB-SAE · concept bottleneck · sparse autoencoders · interpretability · steerability
Official artifacts: Proceedings PDF · Supplement · Author repository (code not yet released) · arXiv
Metadata checked 2026-10-09
2026EMNLPAccepted · proceedings pending
IntroConformal: Conformal Factuality Guarantees for Large Vision-Language Models via Introspective Signals
Md. Atabuzzaman, Christian Alexander, and Chris Thomas
L3
The arXiv v1 record submitted 1 September 2026 reports acceptance to the EMNLP 2026 main conference. No final ACL Anthology record was available when checked on 2 September 2026, so publication claims remain author-reported and scientific claims use the arXiv manuscript.
V&L evaluation / safety LLM evaluation / safety Evaluation / reproducibility
vision–language evaluation / safety · LLM evaluation / safety · evaluation / reproducibility · IntroConformal · conformal risk control · VLM factuality · selective prediction · introspective signal · abstention
Official artifacts: Official implementation
Metadata checked 2026-09-02
2026Scientific ReportsPeer reviewed
KDTViT Knowledge Distillation, Transfer Learning and Transformer Based Deep Learning Framework for Efficient Histopathology Image Classification
Aarv Mankodi, Praveen Kumar Shukla, and Hitesh Tekchandani
L3
Peer-reviewed open-access journal article 16:20721. A MobileViT-XS is sequentially adapted from 40× through 400× using the previous checkpoint as teacher; BreakHis consists of biopsy images rather than WSIs, and the same-capacity teacher/student design targets retention rather than model compression.
Knowledge distillation Transfer / adaptation / continual
knowledge distillation / model transfer · cell / tissue modeling · evaluation · robustness / domain shift / privacy · KDT-ViT · progressive cross-magnification learning · MobileViT · BreakHis · patch classification
Official artifacts: Open-access article
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Keep It Simple: Multi-Key Episodic Memory Retrieval for Ultra-Long Video Understanding
Yeeun Choi, Youngbeom Yoo, Joon-Young Lee, Hyolim Kang, and Seon Joo Kim
L3
ECCV 2026 Oral verified in the official program (event 6047); scientific claims use the conference-hosted manuscript #10831. Publisher proceedings/version of record was not verified on 2026-09-12.
Generation / video / agents Multimodal retrieval
multimodal generation / agents · multimodal embedding / retrieval · video retrieval · episodic memory · temporal expansion
Official artifacts: ECCV program · Conference manuscript · Author project · Code
Metadata checked 2026-09-12
2026ICLRPeer reviewed
Knowledge Distillation for Large Language Models through Residual Learning
Thinh On, Hengzhi Pei, Leonard Lausen, and George Karypis
L3
Published at ICLR 2026; no official implementation was located by 2026-09-02.
Knowledge distillation Transformers / pretraining Generation / MT evaluation LLM systems LLM post-training
knowledge distillation / model transfer · NLP transformers / pretraining · generation / translation evaluation · LLM training / serving systems · LLM post-training / alignment · LLM distillation · residual hidden states · teacher errors · cross-tokenizer distillation · mixture of experts
Official artifacts: OpenReview record · First-party Amazon Science page
Metadata checked 2026-09-02
2026MICCAIPeer reviewed
LaGuadia: Language-Guided Adaptive Distillation from Pathology Foundation Models
Gangsu Kim and Won-Ki Jeong
L3
Official MICCAI 2026 proceedings record and accepted-version open-access PDF verified 2026-10-02; this supersedes the prior preprint classification. Springer DOI remains pending.
Knowledge distillation
knowledge distillation / model transfer · foundation encoder · pathology VLM · MIL / WSI · LaGuadia · language-guided distillation · multi-teacher transfer · pathology foundation models · multimodal student
Official artifacts: Official code · Official MICCAI 2026 proceedings · Official accepted-version PDF
Metadata checked 2026-10-02
2026arXivPreprint
LAION-BVD: A 10-Million-Hour Open Video Dataset for Multimodal Pre-training
Andreas Hochlehnert, Marianna Nezhurina, Mehdi Cherti, Andrej Radonjic, Thaddäus Wiedemer, Christoph Schuhmann, Romain Beaumont, Wieland Brendel, Bernhard Schölkopf, A. Sophia Koepke, Jenia Jitsev, and Matthias Bethge
L2–L3
arXiv preprint first submitted 25 August 2026; not peer reviewed as of the audit date.
Image–text representation Generation / video / agents V&L evaluation / safety Multimodal retrieval
image–text representation · multimodal generation / agents · vision–language evaluation / safety · multimodal embedding / retrieval · LAION-BVD · video dataset · multimodal pretraining · synthetic captions · web-scale data
Official artifacts: Official project · Official data tooling · Official URL dataset
Metadata checked 2026-08-30
2026arXivPreprint
LanGuSTE: Language-Guided Coarse-to-Fine Patch Selection for Efficient Whole Slide Image Analysis
Yonghan Shin, Gangsu Kim, and Won-Ki Jeong
L3
arXiv preprint first submitted in 2025 and substantially revised as v2 on 24 August 2026; no peer-reviewed version or official implementation was verified by the review cutoff.
Transfer / adaptation / continual Image–text representation
MIL / WSI · WSI engineering · pathology VLM · image–text representation · LanGuSTE · coarse-to-fine patch selection · cross-scale visual prompt tuning · LLM-derived concepts · 5× to 20× selection
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
Learning to Deny: Action Denial in Multimodal Large Language Models
Raiyaan Abdullah, Shehreen Azad, and Yogesh Singh Rawat
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #3692 checked on 2026-09-12. Publisher version of record was not independently verified.
V&L evaluation / safety Generation / video / agents Causal inference
vision–language evaluation / safety · multimodal generation / agents · causal inference · UCF101-AD · CausalAct · action absence
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026CVPRPeer reviewed
LensWalk: Agentic Video Understanding by Planning How You See in Videos
Keliang Li et al.
L3
Generation / video / agents
multimodal generation / agents · agentic video · active observation · long video
Metadata checked 2026-07-27
2026arXivPreprint
LeVJEPA: Efficient & Scalable Video Pretraining without the Heuristics
Lukas Kuhn, Lucas Maes, Giuseppe Serra, Quentin Le Lidec, Yann LeCun, Randall Balestriero, and Florian Buettner
L3
arXiv v1 preprint submitted 27 August 2026; compute and transfer claims have not yet received proceedings review.
Generative / representation Visual representations Video / 3D / embodied
generative / representation learning · visual representation learning · video / 3D / embodied vision · LeVJEPA · video pretraining · SIGReg · token dropping · block-causal attention
Official artifacts: Official project · Official implementation · Official model
Metadata checked 2026-08-30
2026CVPRPeer reviewed
LIFT and PLACE: A Simple, Stable, and Effective Knowledge Distillation Framework for Lightweight Diffusion Models
Hyunsoo Han, Sangyeop Yeo, and Jaejun Yoo
L2–L3
Knowledge distillation Generative / representation Generative vision
knowledge distillation / model transfer · generative / representation learning · generative vision · diffusion compression · coarse-to-fine distillation · linear fitting · difficulty-adaptive grouping
Metadata checked 2026-08-10
2026CVPRPeer reviewed
LLaDA-V: Large Language Diffusion Models with Visual Instruction Tuning
Zebin You et al.
L3
Multimodal LLMs
multimodal LLM · diffusion language model · visual instruction · bidirectional attention
Official artifacts: Code
Metadata checked 2026-07-27
2026CVPRPeer reviewed
LLaVAShield: Safeguarding Multimodal Multi-Turn Dialogues in Vision-Language Models
Guolei Huang et al.
L3
V&L evaluation / safety
vision–language evaluation / safety · multi-turn safety · red teaming · cross-modal risk
Metadata checked 2026-07-27
2026ACLPeer reviewed
Look Light, Think Heavy: What Multimodal Chain-of-Thought Reasoning Can and Cannot Do
Zhuoran Jin, Kejian Zhu, Hongbang Yuan, Yupu Hao, Pengfei Cao, Yubo Chen, Kang Liu, and Jun Zhao
L2–L3
Peer-reviewed ACL 2026 long paper.
Captioning / VQA Multimodal LLMs V&L evaluation / safety Reasoning / verification
captioning / VQA · multimodal LLM · vision–language evaluation / safety · LLM reasoning / verification · multimodal chain of thought · visual reasoning · grounding · counting · attention analysis
Official artifacts: Author manuscript
Metadata checked 2026-08-30
2026arXivPreprint
LoopArena: Benchmarking Models as Runtime Controllers for Loop Engineering
Yi Wang, Haopeng Zhang, Chengxiang Huang, Rui Dai, Kaikui Liu, Piotr Koniusz, and Xiangxiang Chu
L3
arXiv v1 submitted 28 August 2026. The exact title follows the current arXiv record, superseding an earlier discovery-feed paraphrase. Benchmark data and evaluation code were released; no peer-reviewed record was verified.
RAG / tools / agents Code / software engineering LLM evaluation / safety
RAG / tools / agents · code generation / software engineering · LLM evaluation / safety · LoopArena · coding agents · runtime controller · controller-worker · loop engineering · software benchmark
Official artifacts: Official benchmark and evaluation code
Metadata checked 2026-09-02
2026MICCAIPeer reviewed
LRMIL: Efficient Low-Resolution Multiple Instance Learning via High-Resolution Knowledge Distillation for Whole Slide Image Classification
Yonghan Shin and Won-Ki Jeong
L3
Official MICCAI 2026 proceedings record and accepted-version open-access PDF verified 2026-10-02. Springer DOI and pagination remain pending; do not invent either.
Knowledge distillation
MIL / WSI · WSI engineering · cross-resolution distillation · low-resolution MIL · efficient WSI inference · knowledge distillation · knowledge distillation / model transfer
Official artifacts: Official code · Official MICCAI 2026 proceedings · Official accepted-version PDF
Metadata checked 2026-10-02
2026CVPRPeer reviewed
M4V: Multimodal Mamba for Efficient Text-to-Video Generation
Jiancheng Huang et al.
L3
Generation / video / agents
multimodal generation / agents · text-to-video · state-space model · efficient generation
Metadata checked 2026-07-28
2026ECCVAccepted · proceedings pending
Make Geometry Matter for Spatial Reasoning
Shihua Zhang, Qiuhong Shen, Shizun Wang, Tianbo Pan, and Xinchao Wang
L3
ECCV 2026 Oral verified in the official program (event 6086); scientific claims use the conference-hosted manuscript #482. Publisher proceedings/version of record was not verified on 2026-09-12.
Captioning / VQA Multimodal LLMs 3D / spatial vision
captioning / VQA · multimodal LLM · 3D / spatial vision · geometry-aware fusion · spatial reasoning · shortcut suppression
Official artifacts: ECCV program · Conference manuscript · Author project · Code
Metadata checked 2026-09-12
2026ICLRPeer reviewed
Mamba-3: Improved Sequence Modeling using State Space Principles
Aakash Sunil Lahoti, Kevin Li, Berlin Chen, Caitlin Wang, Aviv Bick, Zico Kolter, Tri Dao, Albert Gu
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Three selective-SSM changes evaluated on language and synthetic sequences. Distinguish conditional discretization theory from learned gates, parameter matching from compute matching, and fixed-state decoding from information-retrieval capability. No direct pathology or natural-image evaluation.
Text / sequence models Transformers / pretraining LLM systems
text / sequence modeling · NLP transformers / pretraining · LLM training / serving systems · Mamba-3 · state space model · MIMO · complex states · input-dependent rotations · efficient decoding · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview · Official implementation
Metadata checked 2026-09-19
2026ACLPeer reviewed
Massively Multilingual Joint Segmentation and Glossing
Michael Ginn, Lindia Tjuatja, Enora Rice, Ali Marashian, Maria Valentini, Jasmine Xu, Graham Neubig, and Alexis Palmer
L2–L3
Peer-reviewed ACL 2026 long paper and Outstanding Paper award recipient.
Text / sequence models Structure / extraction Multilingual / low resource
text / sequence modeling · linguistic structure / information extraction · multilingual / low-resource NLP · PolyGloss · interlinear glossed text · morphological segmentation · multilingual NLP · low-resource languages
Official artifacts: Author manuscript · Official code · Official models and data
Metadata checked 2026-08-30
2026ACLPeer reviewed
Maximizing Local Entropy Where It Matters: Prefix-Aware Localized LLM Unlearning
Naixin Zhai, Pengyang Shao, Binbin Zheng, Yonghui Yang, Fei Shen, Long Bai, and Xun Yang
L2–L3
Peer-reviewed ACL 2026 long paper and Outstanding Paper award recipient.
LLM interpretability LLM evaluation / safety LLM post-training
LLM interpretability / editing / unlearning · LLM evaluation / safety · LLM post-training / alignment · PALU · machine unlearning · prefix localization · entropy maximization · top-k vocabulary
Official artifacts: Author manuscript · Official code
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
Mechanistic Finetuning of Vision-Language-Action Models via Few-Shot Demonstrations
Chancharik Mitra, Yusen Luo, Raj Saravanan, Dantong Niu, Anirudh Pai, Jesse Thomason, Trevor Darrell, Abrar Anwar, Deva Ramanan, and Roei Herzig
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #13378 checked on 2026-09-12. Publisher version of record was not independently verified.
Generation / video / agents Transfer / adaptation / continual Interpretability
multimodal generation / agents · interpretability / attribution · Robotic Steering · head-selective LoRA · leave-trajectory-out retrieval
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
MIMFlow: Integrating Masked Image Modeling with Normalizing Flows for End-to-End Image Generation
Yang Chen, Xiaowei Xu, Shuai Wang, Xinwen Zhang, Qiushi Guo, Tiezheng Ge, and Limin Wang
L3
ECCV 2026 Oral verified in the official program; scientific claims use the conference-hosted manuscript #2272. Publisher proceedings/version of record was not verified on 2026-09-12.
Generative / representation Generative vision Visual representations
generative / representation learning · generative vision · visual representation learning · MIMFlow · masked image modeling · normalizing flow · learnable latent bottleneck · joint tokenizer training
Official artifacts: ECCV Oral program · Conference manuscript · Official implementation · arXiv manuscript
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Mimicking Radiologists: A Coarse-to-Fine Framework with Structural Sparse Tokens for Dual-LLM Computed Tomography Report Generation
Hong Liu, Dong Wei, Yefeng Zheng, Xian Wu, and Liansheng Wang
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #9339. Publisher proceedings/version of record was not verified on 2026-09-12. The paper-linked CTRG repository README still describes the earlier Structure Observation Driven Image-Text Contrastive Learning work and LLaMA-2-7B. A matching dual-LLM/GRPO release was not verified; do not treat this link as reproduction evidence.
Multimodal LLMs Captioning / VQA Loss / objectives
multimodal LLM · captioning / VQA · loss / objective design · CT report generation · AP-LTF · structure-specific tokens · dual LLM · GRPO
Official artifacts: ECCV Spotlight program · Conference manuscript · Paper-linked repository (earlier method; matching release unverified)
Metadata checked 2026-09-12
2026arXivPreprint
Modular TTT: Rethinking Test-Time Training as Composable Modules
Bohao Tang, Zhen Qin, Yuqi Pan, Zheng Li, Pengfei Liu, and Ya Zhang
L2–L3
Preprint status retained; no peer-reviewed proceedings record was available when checked on 2026-08-10.
Neural networks LLM pretraining LLM systems
neural networks · LLM pretraining / scaling · LLM training / serving systems · Modular TTT · test-time training layers · stateful sequence models · composable modules
Official artifacts: Official code
Metadata checked 2026-08-10
2026ECCVAccepted · proceedings pending
MoE-KD: Your Teacher Model is Worth Mixture-of-Experts for Knowledge Distillation
Kuo Shi, Wenjie Zhu, Bo Peng
L3
ECCV 2026 poster verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. The EM convergence discussion assumes an improving M-step; it is not a guarantee of global optimum for stochastic neural training. Table 3 values take precedence over inconsistent prose deltas. No runnable official implementation was verified in this pass.
Knowledge distillation Loss / objectives Image / recognition
knowledge distillation / model transfer · loss / objective design · image recognition · MoE-KD · latent teacher classes · expectation maximization · teacher classifier reuse · heterogeneous distillation
Official artifacts: Official ECCV poster record
Metadata checked 2026-09-12
2026arXivPreprint
Molmo 2: Open Weights and Open Data for State-of-the-Art Vision-Language Models
Matt Deitke et al.
L2–L3
Multimodal LLMs Grounding / open vocabulary V&L evaluation / safety
multimodal LLM · grounding / open vocabulary · vision–language evaluation / safety · Molmo 2 · video grounding · open data
Official artifacts: Official code
Metadata checked 2026-08-08
2026ECCVAccepted · proceedings pending
MolmoPoint: Better Pointing for VLMs with Grounding Tokens
Christopher Clark, Yue Yang, Jae Sung Park, Zixian Ma, Jieyu Zhang, Rohun Tripathi, Mohammadreza Salehi, Sangho Lee, and Ranjay Krishna
L3
ECCV 2026 Spotlight verified in the official program (event 6028); scientific claims use the conference-hosted manuscript #8311. Publisher proceedings/version of record was not verified on 2026-09-12. The program spells the name Molmo-Point; the conference PDF uses MolmoPoint.
Grounding / open vocabulary Multimodal LLMs Generation / video / agents
grounding / open vocabulary · multimodal LLM · multimodal generation / agents · grounding tokens · coarse-to-fine pointing · video tracking · Molmo-Point: Better Pointing for VLMs with Grounding Tokens
Official artifacts: ECCV program · Conference manuscript · Author project
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Multi-Anchor Distillation with Text-Guided Analytic Classifier for Continual Learning
Qier Meng, De Cheng, Jiahao Li, and Cheng Deng
L3
ECCV 2026 Spotlight verified via the official related presentation event; the poster identity itself says Accept Poster. No Springer version of record or public runnable implementation was verified. Section 3.5's printed closed-form equation omits a term implied by its objective; the review flags this reproducibility issue without silently repairing the authors' formula.
Knowledge distillation Transfer / adaptation / continual Image–text representation Image / recognition
knowledge distillation / model transfer · image–text representation · image recognition · multi-anchor distillation · analytic classifier · class-incremental learning · CLIP · LoRA · exemplar-free · ECCV 2026 spotlight
Official artifacts: Official ECCV 2026 Spotlight
Metadata checked 2026-09-12
2026MICCAI Efficient Medical AI workshop (EMA)Peer reviewed
Multi-Teacher Contrastive Distillation for Edge-Efficient Pathology Foundation Models
Tim Lenz, Maurice Heide, Marco Gustav, Nic G. Reitsam, and Jakob Nikolas Kather
L3
Official MICCAI satellite proceedings confirm the 2nd Efficient Medical AI workshop (EMA), not the main MICCAI track. LNCS volume 17265 is listed; DOI and pagination remain pending on the checked record. Full 11-page accepted manuscript inspected. Teachers and students use 2.0 µm/px (5×); this is capacity compression, not a demonstrated 20×→5× transfer method. Raspberry Pi timings concern individual encoder forward passes, not complete WSI processing.
Knowledge distillation
foundation encoder · self-supervised learning · WSI engineering · knowledge distillation / model transfer · MuCoDi · MuCoEdge · multi-teacher contrastive distillation · edge inference · cached teacher embeddings · same-magnification compression
Official artifacts: Official workshop accepted manuscript · Author implementation and documented scope
Metadata checked 2026-10-26
2026Findings of EMNLPAccepted · proceedings pending
MVC-Bench: Benchmarking Calibration of Medical Vision-Language Models
Ashshak Sharifdeen, Shihab Aaqil Ahamed, Ufaq Khan, Muhammad Akhtar Munir Sujair Ibrahim, Mohamed Rafeek Mareer Ahamed, Yutong Xie, Imran Razzak, and Muhammad Haris Khan
L3
The arXiv record reports acceptance to Findings of EMNLP 2026. No final ACL Anthology record was available when checked on 30 August 2026, so claims were verified against arXiv v1 and the official repository.
V&L evaluation / safety Vision evaluation Evaluation / reproducibility
pathology VLM · evaluation · robustness / domain shift / privacy · vision–language evaluation / safety · vision evaluation / robustness · evaluation / reproducibility · MVC-Bench · medical VLM calibration · histopathology domain shift · expected calibration error · prompt tuning · multi-class margin regularization
Official artifacts: Official implementation
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
NarrativeTrack: Evaluating Entity-Centric Reasoning for Narrative Understanding
Hyeonjeong Ha, Jinjin Ge, Bo Feng, Kaixin Ma, and Gargi Chakraborty
L3
ECCV 2026 Spotlight verified in the official program (event 6177); scientific claims use the conference-hosted manuscript #10972. Publisher proceedings/version of record was not verified on 2026-09-12. Earlier arXiv versions used the subtitle Evaluating Video Language Models Beyond the Frame; this is the same canonical work.
V&L evaluation / safety Generation / video / agents
vision–language evaluation / safety · multimodal generation / agents · entity-centric evaluation · temporal continuity · narrative understanding
Official artifacts: ECCV program · Conference manuscript · Author project · Dataset
Metadata checked 2026-09-12
2026arXivPreprint
Normalized Low-Rank Adaptation
Jiale Kang, Ziyin Yue, Zheng Zhan, Yangyi Huang, and Weiyang Liu
L3
arXiv v1 submitted 31 August 2026. The authors released an implementation; no peer-reviewed record was verified on 2 September 2026.
Transfer / adaptation / continual Transformers / pretraining LLM post-training
NLP transformers / pretraining · LLM post-training / alignment · NoRA · LoRA · normalization · parameter-efficient finetuning · RLVR · catastrophic forgetting
Official artifacts: Official implementation
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Off the Planckian Locus: Using 2D Chromaticity to Improve In-Camera Color
SaiKiran Tedla, Joshua Little, Hakki C. Karaimer, and Michael S. Brown
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #1413. Publisher proceedings/version of record was not verified on 2026-09-12.
Image / recognition Restoration / inverse problems
image recognition · CCM-MLP · color constancy · 2D chromaticity · camera pipeline · LED calibration
Official artifacts: ECCV Spotlight program · Conference manuscript · Author project (paper/program link)
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
OmniForcing: Unleashing Real-time Joint Audio-Visual Generation
Yaofeng Su, Yuming Li, Zeyue Xue, Jie Huang, Siming Fu, Haoran Li, Haoyang Huang, and Nan Duan
L3
ECCV 2026 Oral verified in the official program; scientific claims use the conference-hosted manuscript #9084. Publisher proceedings/version of record was not verified on 2026-09-12. Author list follows the conference manuscript and program; the current repository citation lists two additional authors. Review separates the paper's three-stage method from the later simplified two-stage code release.
Knowledge distillation Generative vision Generation / video / agents
knowledge distillation / model transfer · generative vision · multimodal generation / agents · ECCV 2026 · Oral · audio-visual generation · causal distillation · attention sinks · self-forcing
Official artifacts: ECCV Oral program · Conference manuscript · Official implementation · arXiv manuscript · Author project · Model checkpoint
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
OmniScript: Towards Audio-Visual Script Generation for Long-Form Cinematic Video
Junfu Pu, Yuxin Chen, Teng Wang, and Ying Shan
L3
ECCV 2026 Oral verified in the official program (event 6045); scientific claims use the conference-hosted manuscript #9919. Publisher proceedings/version of record was not verified on 2026-09-12.
Generation / video / agents Captioning / VQA Multimodal LLMs
multimodal generation / agents · captioning / VQA · multimodal LLM · audio-visual scripting · temporal field evaluation · segmented reward
Official artifacts: ECCV program · Conference manuscript · Author project · Code
Metadata checked 2026-09-12
2026CVPRPeer reviewed
OmniVGGT: Omni-Modality Driven Visual Geometry Grounded Transformer
Hengchao Peng et al.
L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · OmniVGGT · multimodal geometry · 3D reconstruction · 3D / spatial vision
Metadata checked 2026-07-27
2026Pattern RecognitionPeer reviewed
On learning denoisable student logits
Diqi Chen, Yang Li, Jiajun Liu, Brano Kusy, Jun Zhou, and Yongsheng Gao
L3
Peer-reviewed article 113684 in Pattern Recognition 179; available online 10 April 2026 and assigned to the November 2026 issue.
Knowledge distillation Generative / representation Image / recognition
knowledge distillation / model transfer · generative / representation learning · image recognition · denoisable logits · diffusion recovery · logit residual · distributed inference · logit compression
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
On Test-Time Scaling for Vision-Language Models
Fawaz Sammani, Tzoulio Chamiti, and Nikos Deligiannis
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #4968 checked on 2026-09-12. Publisher version of record was not independently verified.
V&L evaluation / safety Reasoning / verification Multimodal LLMs
vision–language evaluation / safety · LLM reasoning / verification · multimodal LLM · test-time scaling · visual grounding · token budget
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026arXivPreprint
On-policy Distillation with Verifiable Reward
Wenze Lin, Jiale Zhao, Xitai Jiang, Songde Rao, Yining Li, Shenzhi Wang, Bingxiang He, and Gao Huang
L3
arXiv v1 preprint submitted 25 August 2026; evaluation is limited to the reported verifiable reasoning tasks.
Knowledge distillation AI / reinforcement learning Reasoning / verification LLM post-training
knowledge distillation / model transfer · AI / reinforcement learning · LLM reasoning / verification · LLM post-training / alignment · OPDVR · on-policy distillation · verifiable reward · token credit · reasoning
Official artifacts: Official implementation
Metadata checked 2026-08-30
2026arXivPreprint
On-Policy Self-Distillation in Diffusion Models
Wei Zhou, Xiongwei Zhu, Lingdong Kong, Bo Chen, Lei Zhang, Yongyuan Liang, Xiaoxia Hou, Ye Tian, Xian Sun, Yingshuo Wang, Linfeng Li, Shengqiong Wu, Leigang Qu, Feng Li, Wei Liu, Julian McAuley, and Tat-Seng Chua
L3
arXiv v1 technical report submitted 25 August 2026; reported reward gains rely on the chosen automated evaluators and two tested backbones.
Knowledge distillation Generative / representation Generative vision
knowledge distillation / model transfer · generative / representation learning · generative vision · DiffusionOPSD · on-policy self-distillation · diffusion model · reward gradient · EMA
Official artifacts: Official project · Official implementation
Metadata checked 2026-08-30
2026CVPRPeer reviewed
OneCAT: Decoder-Only Auto-Regressive Model for Unified Understanding and Generation
Han Li et al.
L3
Generation / video / agents
multimodal generation / agents · unified multimodal · autoregressive · understanding and generation
Metadata checked 2026-07-27
2026MIDLPeer reviewed
PAGET: Hierarchical Multi-Teacher Knowledge Distillation for Comprehensive Tumor Microenvironment Segmentation
Daisuke Komura, Maki Takao, Mieko Ochi, Takumi Onoyama, Hiroto Katoh, Hiroyuki Abe, Hiroyuki Sano, Teppei Konishi, Toshio Kumasaka, Tomoyuki Yokose, Yohei Miyagi, Tetsuo Ushiku, and Shumpei Ishikawa
L3
Peer-reviewed MIDL 2026 proceedings, PMLR 315:256–279; official record verified 2026-10-02.
Knowledge distillation
cell / tissue modeling · knowledge distillation / model transfer · PAGET · hierarchical multi-teacher distillation · TME segmentation · IHC restaining · semantic segmentation · panoptic segmentation
Official artifacts: Official proceedings PDF · Official implementation
Metadata checked 2026-10-02
2026arXivPreprint
PANDA: Prototype-Anchored Alignment for Partially Unpaired Multimodal Learning, with Applications to Alzheimer's MRI and TCGA Pathology
Sheethal Bhat, Mahfuzur Rahman Chowdhury, Paula Andrea Perez-Toro, Stephan Wunderlich, Rose Dawn Bharat, Siming Bayer, and Andreas Maier
L3
arXiv v1 preprint submitted 26 August 2026. The manuscript promises code upon acceptance, but no public implementation was available at the review cutoff.
Transfer / adaptation / continual
spatial / molecular · MIL / WSI · evaluation · PANDA alignment · partially paired modalities · class prototypes · WSI–RNA transfer · survival prediction · primary-modality deployment
Metadata checked 2026-08-30
2026Pacific Symposium on BiocomputingPeer reviewed
Patch-level phenotype identification via weakly supervised neuron selection in sparse autoencoders for CLIP-derived pathology embeddings
Keita Tamura, Yao-zhong Zhang, Yohei Okubo, and Seiya Imoto
L3
PSB 2026 volume 31, pages 708–721, with source supplement checked. Neuron discovery uses slide labels, but the reported localization thresholds use patch-level validation annotations. Single-neuron localization does not preserve all multi-feature slide-classification performance.
Interpretability
MIL / WSI · evaluation · interpretability / attribution · SAE · SAE-1N · BatchTopK · weak supervision · CONCH · phenotype localization
Official artifacts: Official PSB proceedings PDF · Author implementation and supplement
Metadata checked 2026-10-09
2026MICCAI MI4MedFM WorkshopAccepted · proceedings pending
PATH-DISSECT: Interpreting Concept Representation and Classifier Reliance in Pathology Foundation Models with Sparse Autoencoders
Mayur Parvatikar, Xingjian Li, and Min Xu
L3
Official MICCAI 2026 Satellite Events page supplies the 11-page camera-ready; Springer DOI and final pages were pending at inspection. This is MI4MedFM workshop work on tissue patches, not main-conference or patient-level WSI outcome validation. Causal claims concern interventions on a frozen encoder plus linear readout, not biological causation.
Interpretability
evaluation · robustness / domain shift / privacy · interpretability / attribution · SAE · PATH-DISSECT · TopK sparse autoencoders · concept ablation · linear probe reliance · UNI · Virchow2 · SPIDER
Official artifacts: Official camera-ready PDF · Author implementation · Workshop OpenReview record
Metadata checked 2026-10-09
2026ECCVAccepted · proceedings pending
Pathryoshka: Compressing Pathology Foundation Models via Multi-Teacher Knowledge Distillation With Nested Embeddings
Christian Grashei, Christian Brechenmacher, Rao Muhammad Umer, Jingsong Liu, Carsten Marr, Peter J. Schüffler, and Ewa Szczurek
L3
ECCV 2026 poster status reverified against the official complete program (poster 4454). No oral or spotlight designation is recorded. A Springer version of record was not verified on 2026-09-12; the existing review retains its declared manuscript revision.
Knowledge distillation
knowledge distillation / model transfer · foundation encoder · evaluation · Pathryoshka · multi-teacher distillation · nested embeddings · foundation-model compression · compact pathology encoder
Official artifacts: Official ECCV record · Official model · Conference manuscript
Metadata checked 2026-09-12
2026arXivPreprint
Pathway-Structured Privileged Distillation for Deployable Computational Pathology
Yongxin Guo, Hao Lu, Onur C. Koyun, Zhengjie Zhu, Muhammet F. Demir, and Metin N. Gurcan
L3
arXiv v1 (2026-06-01) checked 2026-10-02; no conference acceptance verified. Paired RNA is training-only, while WSI is used at deployment. Pathway-memory readouts are not independently validated molecular measurements.
Knowledge distillation
spatial / molecular · MIL / WSI · knowledge distillation / model transfer · MoPE · privileged information · RNA-to-WSI distillation · pathway experts · shared memory · morphology slots
Official artifacts: Versioned manuscript
Metadata checked 2026-10-02
2026arXivPreprint
PAWBench: How Far Are We from Probabilistically Aligned World Modeling?
Yuandong Pu, Le Zhuo, Sayak Paul, Gabriel Jorge Menezes, Avram Đorđević, Shiyang Li, Yifan Zhou, Bin Fu, Wenlong Zhang, Junjun He, Yu Qiao, Yihao Liu, Jingbo Xing, and Xi Chen
L3
arXiv v1 benchmark preprint submitted 27 August 2026; conclusions concern terminal-outcome distributions in its controlled scenarios.
Video / 3D / embodied Vision evaluation
video / 3D / embodied vision · vision evaluation / robustness · PAWBench · world model · probabilistic alignment · calibration · rollout evaluation
Official artifacts: Official project · Official implementation · Official dataset
Metadata checked 2026-08-30
2026CVPRPeer reviewed
PixelDiT: Pixel Diffusion Transformers for Image Generation
Yongsheng Yu et al.
L3
Generative vision
generative vision · PixelDiT · pixel-space diffusion · diffusion transformer
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
PolarAPP: Beyond Polarization Demosaicking for Polarimetric Applications
Yidong Luo, Chenggong Li, Yunfeng Song, Ping Wang, Boxin Shi, Junchao Zhang, and Xin Yuan
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #5713. Publisher proceedings/version of record was not verified on 2026-09-12.
Restoration / inverse problems Loss / objectives Transfer / adaptation / continual
loss / objective design · PolarAPP · polarization demosaicking · meta-learning · task-aware feature alignment
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026ACLPeer reviewed
POLYCHARTQA: Benchmarking Large Vision-Language Models with Multilingual Chart Question Answering
Yichen Xu et al.
L3
Captioning / VQA V&L evaluation / safety Document intelligence
captioning / VQA · vision–language evaluation / safety · multilingual chart QA · low-resource languages · benchmark · document intelligence
Metadata checked 2026-07-28
2026ECCVAccepted · proceedings pending
Poppy: Polarization-based Plug-and-Play Guidance for Enhancing Monocular Normal Estimation
Irene Kim, Sai Tanmay Reddy Chakkera, Alexandros Graikos, Dimitris Samaras, and Akshat Dave
L3
Accepted to ECCV 2026; oral presentation verified in the official program. The conference-hosted manuscript is the review authority; a separate Springer proceedings record was not verified. Title follows the PDF; the program calls the task “Surface Normal Estimation”. Training-free means no backbone retraining, not no test-time optimization.
Restoration / inverse problems 3D / spatial vision
3D / spatial vision · Poppy · polarization guidance · surface normals · test-time optimization · residual image offsets · frozen backbone · Poppy: Polarization-Based Plug-and-Play Guidance for Enhancing Surface Normal Estimation
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · Author project · arXiv record · Official code (inspected, not executed)
Metadata checked 2026-09-12
2026ICLRPeer reviewed
Pre-training under infinite compute
Konwoo Kim, Suhas Kotha, Percy Liang, Tatsunori Hashimoto
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Fixed unique data, regularized teachers and real/synthetic sequence distillation. Infinite-compute claims are fitted asymptotes, not completed runs. Distinguish main mixing recipes from the matched-token ablation and inference parameter savings from total training cost.
Knowledge distillation LLM pretraining
knowledge distillation / model transfer · LLM pretraining / scaling · data-constrained pretraining · sequence distillation · ensemble distillation · self-distillation · real synthetic mixture · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview
Metadata checked 2026-09-19
2026arXivPreprint
Pretraining Multiple Instance Learning Networks with Multi-Teacher Distillation from Pathology Slide Foundation Models
Mingxi Fu, Jiawen Li, Renao Yan, Jiali Hu, Qiehe Sun, Tian Guan, and Yonghong He
L3
arXiv v1 (2026-07-16) checked 2026-10-02. Six cohorts cover 15 tasks, not 15 independent datasets. Official code and pretrained aggregator checkpoints are listed; experiments were not rerun. This transfers slide embeddings into MIL initialization, not 20×-to-5× acquisition.
Knowledge distillation Loss / objectives
MIL / WSI · foundation encoder · self-supervised learning · knowledge distillation / model transfer · loss / objective design · multi-teacher distillation · MIL pretraining · ADN · TITAN · CARE · aggregator initialization
Official artifacts: Versioned manuscript · Official implementation and pretrained MIL weights
Metadata checked 2026-10-02
2026ECCVAccepted · proceedings pending
PRISM-VO: Scale-Aware Visual Odometry Using Photometric Plenoptic Bundle Adjustment
Aymeric Fleith, Julian Zirbel, Daniel Cremers, and Niclas Zeller
L3
Accepted to ECCV 2026; oral presentation verified in the official program. The conference-hosted manuscript is the review authority; a separate Springer proceedings record was not verified. The official project announces forthcoming code; a runnable public release was not verified.
3D / spatial vision Video / 3D / embodied
3D / spatial vision · video / 3D / embodied vision · PRISM-VO · plenoptic camera · photometric bundle adjustment · metric scale · uncertainty weighting
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · Author project · arXiv record
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Provable and Robust Wavefront Sensing via Self-Reference Interferometry
Nebiyou Yismaw, Vishwanath Saragadam, Aswin C. Sankaranarayanan, and M. Salman Asif
L3
Accepted to ECCV 2026; oral presentation verified in the official program. The conference-hosted manuscript is the review authority; a separate Springer proceedings record was not verified. Official implementation inspected, not executed; its README labels the sequential v2 least-squares tail unfinished and uses the vectorized v3 path.
Restoration / inverse problems
coprime PSI · phase retrieval · wavefront sensing · self-reference interferometry · graph propagation
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · Author project · arXiv record · Official code (inspected, not executed)
Metadata checked 2026-09-12
2026arXivPreprint
Puro-2B: Poor Lab's Qwen2-1.5B Trained on RTX 5090 within $5090
Kairong Luo, Jiarui Cui, Yaorui Yin, Shengqi Chen, Yiming Yang, Linxiang Gao, Yanmohan Wang, Mingzhe Zhang, Kaiyue Wen, Kaifeng Lyu, and Wenguang Chen
L3
arXiv v1 submitted 27 August 2026. Catalog authors follow the current arXiv metadata; the PDF title page additionally names Chengxia Li, an unresolved source-metadata discrepancy that should be rechecked on revision. Code, data, and checkpoints were released, but no peer-reviewed record was verified.
LLM pretraining LLM systems Evaluation / reproducibility
LLM pretraining / scaling · LLM training / serving systems · evaluation / reproducibility · Puro-2B · open pretraining · RTX 5090 · FP8 · MuonH · cost scaling
Official artifacts: Official training implementation · Official models and data collection
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Pushing the Limits of High-Resolution Weather Forecasting through Data Scaling
Yang Zhao, Peisong Niu, Tian Zhou, Ziqing Ma, Guanlong Ma, Rong Jin, Huiling Yuan, and Liang Sun
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #7258. Publisher proceedings/version of record was not verified on 2026-09-12.
Restoration / inverse problems Generative / representation Evaluation / reproducibility
generative / representation learning · evaluation / reproducibility · BaguanHR · weather forecasting · super-resolution data scaling · self-adaptive rollout
Official artifacts: ECCV Spotlight program · Conference manuscript · Author project (paper/program link)
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
PuzLM: Solving Jigsaw Puzzles with Sequence-to-Sequence Language Models
Gur Elkin, Ofir Itzhak Shahar, and Ohad Ben-Shahar
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #3026 checked on 2026-09-12. Publisher version of record was not independently verified.
3D / spatial vision Text / sequence models
3D / spatial vision · text / sequence modeling · jigsaw reconstruction · discrete visual tokenization · encoder-decoder
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026arXivPreprint
Qwen3-VL-Embedding: Advancing Multimodal Retrieval and Reranking
Qwen Team
L2–L3
Multimodal retrieval Multimodal LLMs
multimodal embedding / retrieval · multimodal LLM · Qwen3-VL-Embedding · multimodal retrieval · reranking
Official artifacts: Official code · Official project article
Metadata checked 2026-08-08
2026arXivPreprint
R2M-Bench: Evaluating Revisit Memory via Relative Consistency in Interactive Video World Models
Qiwen Gu, Bingjie Gao, Rui Chen, Geng Li, Jifan Li, Qishuai Wen, Li Niu, Jing Tang, Xiangxiang Chu, and Junqiao Zhao
L3
arXiv v1 benchmark preprint submitted 27 August 2026; the proposed metrics measure observable revisit consistency, not direct access to internal memory state.
Video / 3D / embodied Vision evaluation
video / 3D / embodied vision · vision evaluation / robustness · R2M-Bench · interactive video world model · revisit memory · relative consistency · navigation
Metadata checked 2026-08-30
2026Medical Image AnalysisAccepted · proceedings pending
RACR-MIL: Rank-aware contextual reasoning for weakly supervised grading of squamous cell carcinoma using whole slide images
Anirudh Choudhary, Mosbah Aouad, Krishnakant Saboo, Angelina Hwang, Jacob Kechter, Blake Bordeaux, Puneet Bhullar, David DiCaudo, Steven Nelson, Nneka Comfere, Emma Johnson, Olayemi Sokumbi, Jason Sluzevich, Leah Swanson, Dennis Murphree, Aaron Mangold, and Ravishankar Iyer
L3
The complete arXiv v3 manuscript identifies Medical Image Analysis as the journal and the abstract record reports acceptance; a final DOI or journal issue record was not yet available when checked on 30 August 2026.
Loss / objectives
MIL / WSI · evaluation · loss / objective design · RACR-MIL · ordinal grading · hybrid WSI graph · rank-aware attention · squamous cell carcinoma · pathologist reader study
Metadata checked 2026-08-30
2026CVPRPeer reviewed
RealUnify: Do Unified Models Truly Benefit from Unification? A Comprehensive Benchmark
Yang Shi et al.
L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · unified model · synergy benchmark · understanding and generation
Metadata checked 2026-07-27
2026ICLRPeer reviewed
Reasoning as Representation: Rethinking Visual Reinforcement Learning in Image Quality Assessment
Shijie Zhao, Xuanyu Zhang, Weiqi Li, Junlin Li, Li Zhang, Tianfan Xue, Jian Zhang
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Teacher-generated quality descriptions align a compact image encoder; deployed scoring uses PCA and fitted bases without an LLM. Keep RALI separate from multi-domain RACT. Dataset-specific trade-offs, causal-interpretation limits and the conflicting memory-reduction claim are qualified. No WSI or cross-magnification evaluation.
Knowledge distillation Image–text representation Vision evaluation
knowledge distillation / model transfer · image–text representation · vision evaluation / robustness · RALI · RACT · reasoning distillation · image quality assessment · description alignment · cross-architecture transfer · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview · Official implementation
Metadata checked 2026-09-19
2026ECCVAccepted · proceedings pending
Reconstructing Humans and Objects in Interaction using Large Reconstruction Models
Agniv Chatterjee and Georgios Pavlakos
L3
ECCV 2026 poster presentation is now verified against the official conference inventory (poster 3826), superseding the earlier author-only acceptance evidence. A Springer version of record was not verified on 2026-09-12; the existing review remains tied to its declared manuscript revision.
3D / spatial vision Video / 3D / embodied
3D / spatial vision · video / 3D / embodied vision · MILO · human object interaction · 3D reconstruction · large reconstruction model · SMPL-H
Official artifacts: Official project · Official implementation · Official ECCV 2026 poster
Metadata checked 2026-09-12
2026CVPRPeer reviewed
ReFAct: Empowering Multimodal Web Agents with Visual and Context Focusing
Rui Wu et al.
L3
Generation / video / agents
multimodal generation / agents · web agent · visual focus · context selection
Metadata checked 2026-07-27
2026arXivPreprint
ReFlowSET: Representation-Aligned Latent Flow Matching for SAR-to-EO Image Translation
Jeonghyeok Do, Seungchul Lee, and Munchurl Kim
L3
arXiv v1 submitted 1 September 2026. The exact title follows the current arXiv record, superseding an earlier feed title. Code, weights, and an official project page were available; no peer-reviewed record was verified.
Generative / representation Loss / objectives Generative vision Restoration / inverse problems
generative / representation learning · loss / objective design · generative vision · ReFlowSET · flow matching · SAR-to-EO translation · latent codec audit · conditional DiT · representation alignment
Official artifacts: Official implementation and checkpoints · Official project page
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Register Any Point: Scaling 3D Point Cloud Registration by Flow Matching
Yue Pan, Tao Sun, Liyuan Zhu, Lucas Nunes, Iro Armeni, Jens Behley, and Cyrill Stachniss
L3
Accepted to ECCV 2026; oral presentation verified in the official program. This record and review use the conference-hosted manuscript checked 2026-09-12; a separate Springer proceedings record was not verified. The arXiv preprint first appeared in 2025; the catalog year follows its verified 2026 conference acceptance.
3D / spatial vision Generative / representation
3D / spatial vision · generative / representation learning · RAP · flow matching · point cloud registration · rigidity enforcement · cross-domain registration · multi-view geometry
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · arXiv record · Official code · Official project
Metadata checked 2026-09-12
2026arXivPreprint
Reliable Benchmarking of Artifact Detection in Computational Pathology: A Reproducibility and Uncertainty Analysis
Konstantinos Moutselos and Ilias Maglogiannis
L3
arXiv v1 submitted 31 August 2026. The authors released reproducibility deposits on Zenodo; no peer-reviewed record was verified on 2 September 2026.
Evaluation / reproducibility
WSI engineering · evaluation · robustness / domain shift / privacy · evaluation / reproducibility · artifact detection · WSI quality control · reproducibility · bootstrap uncertainty · partition sensitivity · preprocessing confounding
Official artifacts: Official reproducibility deposit · Referenced benchmark artifact deposit
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Residual-Guided Expert Specialization for Incomplete Multimodal Learning
Seunghun Baek, Jihwan Park, Jaeyoon Sim, Minjae Jeong, Hoseok Lee, Won Hwa Kim
L3
ECCV 2026 poster verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. MARS distills a privileged residual router into an available-feature router; it is not a cascade of residual-error student networks. Full modalities are assumed available during training. Food-101 full-modality accuracy is below SimMLM despite higher missing-modality and average scores.
Knowledge distillation Neural networks Image / recognition Detection / segmentation Image–text representation
knowledge distillation / model transfer · neural networks · image recognition · detection / segmentation · image–text representation · MARS residual routing · missing modality · privileged information · router distillation · mixture of experts · residual-guided specialization
Official artifacts: Official ECCV poster record · Official code
Metadata checked 2026-09-12
2026arXivPreprint
Restoring Without Forgetting: Continual Learning Across Image Degradations
Alif Ashrafee and Bartosz Krawczyk
L3
arXiv v1 preprint submitted 24 August 2026; validation covers the reported five degradation domains rather than arbitrary composite corruption.
Restoration / inverse problems Transfer / adaptation / continual
continual restoration · low-rank adapter · prototype router · catastrophic forgetting · image degradation
Official artifacts: Official implementation
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
RoMa v2: Harder Better Faster Denser Feature Matching
Johan Edstedt, David Nordström, Yushan Zhang, Georg Bökman, Jonathan Astermark, Anders Heyden, Viktor Larsson, Mårten Wadenbäck, Michael Felsberg, and Fredrik Kahl
L3
Accepted to ECCV 2026; spotlight presentation verified in the official program. This record and review use the conference-hosted manuscript checked 2026-09-12; a separate Springer proceedings record was not verified. The first preprint is from 2025; 2026 reflects the verified conference. Author order and table values follow the ECCV-hosted manuscript, not the earlier repository BibTeX.
3D / spatial vision Visual representations
3D / spatial vision · visual representation learning · RoMa v2 · dense feature matching · DINOv3 · predictive covariance · coarse-to-fine correspondence · visual localization
Official artifacts: Official ECCV spotlight presentation · Official ECCV paper record · arXiv record · Official code
Metadata checked 2026-09-12
2026CVPRPeer reviewed
ROSE: Rotate Your Large Language Model to See
Tongtian Yue et al.
L2
Multimodal LLMs
multimodal LLM · parameter rotation · visual injection
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
RPM-Distill: Physiology-guided Adaptive Cross-modal Distillation for Robust Remote Physiological Measurement
Jiyao Wang, Qingyong Hu, Duoxun Tang, Xiao Yang, Kaishun Wu, and Jiangbo Yu
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #10941. Publisher proceedings/version of record was not verified on 2026-09-12.
Knowledge distillation Loss / objectives Video / 3D / embodied Vision evaluation
knowledge distillation / model transfer · loss / objective design · video / 3D / embodied vision · vision evaluation / robustness · RPM-Distill · RF-to-RGB distillation · spectral relation map · bilevel meta-policy · privileged modality
Official artifacts: ECCV Spotlight program · Conference manuscript · Official implementation
Metadata checked 2026-09-12
2026NeurIPSAccepted · proceedings pending
SAE++: Cascaded Sparse Autoencoders Learn Multi-Level Visual Concepts in Multimodal LLMs
Yusong Zhao, Hengyi Wang, Tanuja Ganu, Akshay Nambi, and Hao Wang
L3
NeurIPS 2026 Accept (poster) verified from official program event 151133, OpenReview cl4vXMku1D, matching all five authors. Program title omits SAE++ and joins LearnMulti-Level; review uses arXiv v3 revised 7 October 2026. Proceedings version not yet verified.
Multimodal LLMs Image–text representation V&L evaluation / safety Interpretability
multimodal LLM · image–text representation · vision–language evaluation / safety · interpretability / attribution · SAE · SAE++ · cascaded sparse autoencoders · hierarchical concepts · decoder atoms · group-level steering
Official artifacts: Reviewed manuscript v3 · Official NeurIPS poster · OpenReview · Code
Metadata checked 2026-10-09
2026arXivPreprint
SCOPD: Sparse-Context On-Policy Self-Distillation for Efficient Vision-Language Models
Ahmadreza Jeddi, Enming Zhang, Jasper Gerigk, Hakki Karaimer, Mozhgan Nasr Azadani, Jiayun Luo, Minh Ngoc Le, Gholamali Aminian, Hugo Buurmeijer, Yongchao Chen, Leonid Sigal, Igor Gilitschenski, Konstantinos G. Derpanis, Marco Pavone, and Babak Taati
L3
Preprint checked 2026-10-02; project page still labels code and models as forthcoming. Visual-token retention is not an optical-magnification intervention.
Knowledge distillation Multimodal LLMs Captioning / VQA V&L evaluation / safety
knowledge distillation / model transfer · multimodal LLM · captioning / VQA · vision–language evaluation / safety · SCOPD · privileged visual context · on-policy self-distillation · visual token pruning · visual sensitivity
Official artifacts: Versioned full text · Author project page
Metadata checked 2026-10-02
2026Findings of ACLPeer reviewed
SecureWebArena: A Holistic Security Evaluation Benchmark for LVLM-based Web Agents
Zonghao Ying et al.
L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · web agents · adversarial environments · security evaluation
Metadata checked 2026-07-28
2026ECCVAccepted · proceedings pending
Seeing Through the Weights: Privacy Leakage in Scene Coordinate Regression
Oleksii Nasypanyi, Jaemin Cho, Utku Ozbulak, Byungkon Kang, and Francois Rameau
L3
Accepted to ECCV 2026; oral presentation verified in the official program. This record and review use the conference-hosted manuscript checked 2026-09-12; a separate Springer proceedings record was not verified. The official repository was a README with a forthcoming-code notice when checked; a runnable implementation was not verified.
3D / spatial vision Privacy / federated
3D / spatial vision · privacy / federated learning · scene coordinate regression · model inversion · privacy leakage · visual localization · proxy queries · threat-model evaluation
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · arXiv record · Official project · Official code repository (release announced, not yet implemented)
Metadata checked 2026-09-12
2026arXivPreprint
Self-OPD: On-Policy Distillation for Flow Matching Models without Teacher
Shiyi Zhang, Mushui Liu, Yunze Tong, Wanggui He, Siyu Zou, Jinlong Liu, Yunlong Yu, Jian Song, Hao Jiang, Pipei Huang, and Bo Zheng
L3
arXiv v1 preprint submitted 27 August 2026; teacher-free reward optimization remains benchmark- and reward-model-dependent.
Knowledge distillation Generative / representation Generative vision
knowledge distillation / model transfer · generative / representation learning · generative vision · Self-OPD · flow matching · on-policy distillation · teacher-free · reward optimization
Official artifacts: Official implementation
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
SFKD: Spatial–Frequency Joint-Aware Heterogeneous Knowledge Distillation via Multi-Level Wavelet Spectral Interaction
Cuipeng Wang, Haipeng Wang
L3
ECCV 2026 poster verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. The linked official repository still says source code will be released soon. This is heterogeneous-architecture feature KD, not demonstrated cross-magnification transfer. The selected CIFAR-100 table includes a ViT-S→MobileNetV2 case where FBT is better.
Knowledge distillation Loss / objectives Visual representations Image / recognition
knowledge distillation / model transfer · loss / objective design · visual representation learning · image recognition · SFKD · heterogeneous knowledge distillation · Haar wavelets · Fourier amplitude · spatial-frequency alignment · residual refinement
Official artifacts: Official ECCV poster record · Official code
Metadata checked 2026-09-12
2026ICMLPeer reviewed
Similarity Is Not Logic: Factored Inference for Dual-Encoder Vision-Language Models
Sultan Alshehri, Zhantao Yang, Han Zhang, and Marios Savvides
L3
Multimodal retrieval V&L evaluation / safety Image–text representation
multimodal embedding / retrieval · vision–language evaluation / safety · image–text representation · LCSE · factored inference · visual negation · compositional retrieval · FACTOR-Bench · frozen encoders
Official artifacts: Proceedings PDF · Author project
Metadata checked 2026-10-12
2026SIGGRAPH AsiaAccepted · proceedings pending
SiZeUp: Fast 3D Proxy from Aerial Images via Depth Ordinal Loss
Wenjun Zhou, Yunshan Li, Qiaoyu Zhu, Weidan Xiong, Hao Zhang, Daniel Cohen-Or, and Hui Huang
L3
The arXiv manuscript reports SIGGRAPH Asia 2026 acceptance. No final ACM proceedings record or official implementation was available by 2026-08-30.
3D / spatial vision
3D / spatial vision · SiZeUp · 3D proxy · aerial imagery · ordinal depth loss · building reconstruction
Metadata checked 2026-08-30
2026ICLRPeer reviewed
SliderQuant: Accurate Post-Training Quantization for LLMs
Shigeng Wang, Chao Li, Yangyuxuan Kang, Jiawei Fan, Zhonghong Ou, and Anbang Yao
L3
Peer-reviewed ICLR 2026 poster; retain as a frontier quantization method after systems and quantization foundations.
LLM evaluation / safety LLM systems
LLM evaluation / safety · SliderQuant · post-training quantization · model compression · LLM training / serving systems
Official artifacts: Code
Metadata checked 2026-07-28
2026arXivPreprint
Sliding-window beats linear attention
Alexia Jolicoeur-Martineau, Rhea Sanjay Sukthanker, Pashmina Cameron, and Emy Gervais
L3
arXiv v1 submitted 28 August 2026. No peer-reviewed record, code release, or model artifact was verified on 2 September 2026; it is retained for its unusually important negative baseline evidence rather than publication status.
Transformers / pretraining LLM pretraining LLM systems RAG / tools / agents Evaluation / reproducibility
NLP transformers / pretraining · LLM pretraining / scaling · LLM training / serving systems · RAG / tools / agents · evaluation / reproducibility · sliding-window attention · attention sinks · linear attention · training-free retrofit · long-context evaluation · negative baseline
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
SOCO: Benchmarking Semantic Object Correspondence in Vision Foundation Models
Olaf Dünkel, Basavaraj Sunagad, Haoran Wang, David T. Hoffmann, Christian Theobalt, and Adam Kortylewski
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #349 checked on 2026-09-12. Publisher version of record was not independently verified.
Vision evaluation Visual representations Grounding / open vocabulary V&L evaluation / safety
vision evaluation / robustness · visual representation learning · grounding / open vocabulary · vision–language evaluation / safety · semantic object correspondence · part identity · cross-category evaluation
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Steerable Visual Representations
Jona Ruthardt, Manu Gaur, Deva Ramanan, Makarand Tapaswi, and Yuki M. Asano
L3
ECCV 2026 Oral verified in the official program; scientific claims use the conference-hosted manuscript #14503. Publisher proceedings/version of record was not verified on 2026-09-12. The official program title is 'Steerable Vision Transformers'; the conference-hosted manuscript, arXiv and author project use 'Steerable Visual Representations'.
Visual representations Image–text representation Grounding / open vocabulary Multimodal retrieval
visual representation learning · image–text representation · grounding / open vocabulary · multimodal embedding / retrieval · SteerViT · Steerable Vision Transformers · early fusion · gated cross-attention · prompt-conditioned visual features
Official artifacts: ECCV Oral program · Conference manuscript · Official implementation · arXiv manuscript · Author project · Model checkpoints
Metadata checked 2026-09-12
2026ECCVAccepted · proceedings pending
Stokes-Informed Diffusion for Robust Linear Polarization Estimation
Yidong Luo, Chenggong Li, Yuchao Feng, Boxin Shi, Junchao Zhang, and Xin Yuan
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #5749. Publisher proceedings/version of record was not verified on 2026-09-12.
Restoration / inverse problems Knowledge distillation Loss / objectives
knowledge distillation / model transfer · loss / objective design · GenPolar · Stokes parameters · distribution matching distillation · LoRA · polarization
Official artifacts: ECCV Spotlight program · Conference manuscript
Metadata checked 2026-09-12
2026UAIPeer reviewed
Stop Probing, Start Coding: Why Linear Probes and Sparse Autoencoders Fail at Compositional Generalization
Vitória Barin-Pacela, Shruti Joshi, Isabela Camacho, Simon Lacoste-Julien, and David Klindt
L3
Published in Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337.
Interpretability Evaluation / reproducibility
interpretability / attribution · evaluation / reproducibility · sparse coding · sparse autoencoder · linear probing · compositional generalization · amortization gap
Metadata checked 2026-08-30
2026Medical Image AnalysisPeer reviewed
Streamline Pathology Foundation Model by Cross-Magnification Distillation
Ziyu Su, Abdul Rehman Akbar, Usama Sajjad, Anil V. Parwani, and Muhammad Khalid Khan Niazi
L3
Journal record verified through publisher metadata, Crossref and PubMed PMID 42874158 on 26 October 2026: electronic publication 10 October 2026, assigned to volume 115 (January 2027), article 104352. The stable ID retains the 2025 preprint year. Journal full text was unavailable; detailed scientific claims and visuals remain explicitly tied to the previously inspected arXiv v3, not asserted to match the journal version. The journal abstract describes seven tasks and encoder-level throughput, not an end-to-end speedup.
Knowledge distillation
knowledge distillation / model transfer · foundation encoder · self-supervised learning · WSI engineering · evaluation · XMAG · cross-magnification distillation · UNI2 teacher · compact pathology encoder · 5× inference
Official artifacts: Inspected author manuscript (v3) · PubMed journal metadata and abstract · Official code · Official model
Metadata checked 2026-10-26
2026CVPRPeer reviewed
Streamlined Knowledge Distillation
Hyeon-Jin Jeong, Han-Jin Lee, and Seok-Hwan Choi
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · direction-wise knowledge · Gramian matrix · Mahalanobis distance · logit distillation
Official artifacts: Official code
Metadata checked 2026-08-10
2026ECCVAccepted · proceedings pending
SurvMILKD: A Weakly Supervised Survival Analysis Framework for Multi-Teacher Knowledge Distillation using Pathology Foundation Models
Mayur Mallya, Ali Khajegili Mirabadi, Hossein Farahani, and Ali Bashashati
L3
The official ECCV 2026 poster and conference-hosted manuscript verify acceptance; a Springer version of record was not verified. The official repository was a code-coming-soon placeholder on 2026-09-12. All foundation encoders remain frozen: task-specific MIL adapters and translation modules learn, not the UNI backbone.
Knowledge distillation
MIL / WSI · foundation encoder · evaluation · knowledge distillation / model transfer · SurvMILKD · survival analysis · multi-teacher distillation · risk-aware distillation · frozen encoders · ABMIL · ECCV 2026 poster
Official artifacts: Official ECCV 2026 Poster · Official repository — code pending
Metadata checked 2026-09-12
2026ICLRPeer reviewed
TD-JEPA: Latent-predictive Representations for Zero-Shot Reinforcement Learning
Marco Bagatella, Matteo Pirotta, Ahmed Touati, Alessandro Lazaric, Andrea Tirinzoni
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Reward-free state–action transitions, not action-free video. Zero-shot policy selection still requires reward observations for task-vector inference. Theoretical assumptions, benchmark-specific actor inputs and mixed OGBench comparisons are qualified in the source-checked review; no WSI evaluation.
Generative / representation AI / reinforcement learning Visual representations Video / 3D / embodied
generative / representation learning · AI / reinforcement learning · visual representation learning · video / 3D / embodied vision · TD-JEPA · non-contrastive SSL · temporal difference · zero-shot reinforcement learning · latent prediction · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview · Official implementation
Metadata checked 2026-09-19
2026ACLPeer reviewed
Textual Steering Vectors Can Improve Visual Understanding in Multimodal Large Language Models
Woody Haosheng Gan, Deqing Fu, Julian Asilis, Ollie Liu, Vatsal Sharan, Robin Jia, and Willie Neiswanger
L2–L3
Peer-reviewed ACL 2026 long paper; catalog authors follow the final proceedings record.
LLM interpretability Multimodal LLMs V&L evaluation / safety
LLM interpretability / editing / unlearning · multimodal LLM · vision–language evaluation / safety · textual steering · activation steering · sparse autoencoders · spatial reasoning · multimodal transfer
Official artifacts: Earlier author manuscript
Metadata checked 2026-08-30
2026arXivPreprint
The Devil is in the Spectrum Bias: Spectrum-Balanced Feature Matching for Robust Representation Distillation
Kuniaki Saito and Yoshitaka Ushiku
L3
arXiv v1 checked 2 October 2026. No official code/model release or conference acceptance verified. Camelyon17 is patch-level evaluation, not whole-slide cross-magnification validation. PE teacher/student labels and parameter counts differ between main text/tables and Appendix C; this review uses unambiguous DINOv2-to-MobileNet Table 1 rows.
Knowledge distillation Loss / objectives Image / recognition Visual representations
knowledge distillation / model transfer · loss / objective design · image recognition · visual representation learning · pearson-pca · dinov2 · knowledge-distillation
Metadata checked 2026-10-02
2026CVPRPeer reviewed
The Geometry of Robustness: Optimizing Loss Landscape Curvature and Feature Manifold Alignment for Robust Finetuning of Vision-Language Models
Shivang Chopra, Shaunak Halbe, Chengyue Huang, Brisa Maneechotesuwan, and Zsolt Kira
L3
Vision evaluation Loss / objectives
vision evaluation / robustness · robust fine-tuning · vision-language models · feature alignment · loss / objective design
Metadata checked 2026-07-28
2026arXivPreprint
The Mask Is Not the Model: Auditing Prefix Invariance in Attention, State-Space, and Hybrid Sequence Models
Taebong Kim, Youngsik Hong, Minsik Kim, Sunyoung Choi, Jaewon Jang, and Minseo Kim
L2–L3
arXiv preprint first submitted 24 August 2026; not peer reviewed as of the audit date.
Transformers / pretraining LLM systems LLM evaluation / safety
NLP transformers / pretraining · LLM training / serving systems · LLM evaluation / safety · prefix invariance · causal audit · state-space model · hybrid sequence model · AX-RAY
Official artifacts: Official AX-RAY demo
Metadata checked 2026-08-30
2026ACL System DemonstrationsPeer reviewed
The olmOCR Project: Building Fully Open OCR using VLMs
Jake Poznanski, Kyle Lo, and Luca Soldaini
L2–L3
Peer-reviewed ACL 2026 system demonstration and Best Demonstration Paper award recipient.
Document intelligence V&L evaluation / safety Multimodal LLMs
document intelligence · vision–language evaluation / safety · multimodal LLM · olmOCR · document OCR · vision-language model · visual unit tests · open system
Official artifacts: Official implementation and models
Metadata checked 2026-08-30
2026arXivPreprint
The Safeguard Worked. Is the LLM System Safer?
Pingyu Wu, Weiming Zhang, and Nenghai Yu
L3
arXiv v1 submitted 1 September 2026. No peer-reviewed record or first-party machine-readable evidence artifact was verified; it is retained because its 198-paper evidence audit supplies unusually important negative system-level evidence.
LLM evaluation / safety Evaluation / reproducibility
LLM evaluation / safety · evaluation / reproducibility · LLM safeguards · deployment safety · residual harmful assistance · system evidence · adaptive attacker · systematic evidence audit
Metadata checked 2026-09-02
2026CVPRPeer reviewed
Thinking With Videos: Multimodal Tool-Augmented Reinforcement Learning for Long Video Reasoning
Haoji Zhang et al.
L3
Generation / video / agents
multimodal generation / agents · VITAL · active frame sampling · tool-augmented reasoning
Metadata checked 2026-07-27
2026CVPRPeer reviewed
TIPSv2: Advancing Vision-Language Pretraining with Enhanced Patch-Text Alignment
Bingyi Cao et al.
L3
Image–text representation Grounding / open vocabulary
image–text representation · grounding / open vocabulary · patch-text alignment · distillation · dense transfer
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
Token-level Response-visual Attention Guidance for Multimodal LLMs Knowledge Distillation
Jaehyun Jang, Eunseop Yoon, Hee Suk Yoon, SooHwan Eom, Mark A. Hasegawa-Johnson, and Chang D. Yoo
L3
ECCV 2026 poster verified from the official linked program event. Review uses the conference-hosted manuscript checked 2026-09-12; a separate Springer version of record was not verified. Teacher and student share a frozen SigLIP vision encoder and resolution; this is not cross-resolution KD. The official code repository lists training/evaluation scripts as planned. Attention similarity correlations are not causal guarantees of faithful reasoning.
Knowledge distillation Loss / objectives Multimodal LLMs Captioning / VQA
knowledge distillation / model transfer · loss / objective design · multimodal LLM · captioning / VQA · TRAG · response-to-vision attention · entropy-adaptive KL · MLLM distillation · teacher forcing · visual grounding
Official artifacts: Official ECCV poster record · Official code
Metadata checked 2026-09-12
2026arXivPreprint
Too much of a good thing -- when knowledge distillation promotes overfitting, and how to avoid it
Irene Trigueros-Lorca, Leonardo Concepción, Christian Wagner, Isaac Triguero, and Daniel Molina
L3
arXiv v1 preprint submitted 24 August 2026; conclusions are scoped to the reported homogeneous CNN image-classification study.
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · blockwise distillation · overfitting · intermediate features · data scarcity · image classification
Metadata checked 2026-08-30
2026ICLRPeer reviewed
TRACE: Your Diffusion Model Is Secretly an Instance Edge Detector
Sanghyun Jo, Ziseok Lee, Wooyeol Lee, Jonghyun Choi, Jaesik Park, Kyungsu Kim
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Diffusion-boundary self-distillation improves selected natural-image segmentation but degrades MoNuSeg/TNBC cell segmentation. One-pass speed is edge extraction only, with the large backbone retained. No WSI or cross-magnification validation; official code/checkpoints remain promised rather than released at the checked revision.
Knowledge distillation Detection / segmentation Visual representations
knowledge distillation / model transfer · detection / segmentation · visual representation learning · TRACE · ICLR 2026 oral · self-distillation · diffusion attention · instance edges · negative transfer · histology domain shift · cell / tissue modeling
Official artifacts: OpenReview · Official project · Official repository — release pending
Metadata checked 2026-09-19
2026arXivPreprint
Two Global Crops Suffice: Locating Semantic Emergence in DINO-Style Self-Supervised Learning
Basavaraj Sunagad, Artur Jesslen, and Adam Kortylewski
L3
arXiv:2609.28187v1, submitted 23 September 2026, inspected 26 October 2026. Empirical DINO-family ablations on ImageNet-1k, not a universal necessity theorem or an evaluation of DINOv3/pathology. The review retains a discrepancy between the main text’s fixed-scale depth claim and Appendix Table A3.
Generative / representation Evaluation / reproducibility Visual representations Vision evaluation
generative / representation learning · evaluation / reproducibility · visual representation learning · vision evaluation / robustness · DINO · iBOT · self-supervised learning · global crops · semantic correspondence · SPair-71k · NAVI · controlled ablation · evaluation proxies
Official artifacts: Inspected manuscript and appendices (v1)
Metadata checked 2026-10-26
2026CVPRPeer reviewed
U²Flow: Uncertainty-Aware Unsupervised Optical Flow Estimation
Xunpei Sun, Wenwei Lin, Yi Chang, and Gang Chen
L3
Video / 3D / embodied
video / 3D / embodied vision · optical flow · uncertainty estimation · unsupervised learning
Official artifacts: Code
Metadata checked 2026-07-28
2026CVPRPeer reviewed
Uncertainty-Aware Knowledge Distillation for Multimodal Large Language Models
Jingchen Sun, Shaobo Han, Deep Patel, Wataru Kohno, Can Jin, and Changyou Chen
L2–L3
Knowledge distillation Multimodal LLMs LLM evaluation / safety
knowledge distillation / model transfer · multimodal LLM · LLM evaluation / safety · multimodal LLM compression · uncertainty weighting · energy-based prior · loss balancing
Metadata checked 2026-08-10
2026ECCVAccepted · proceedings pending
Understanding Geometric Representations in Self-Supervised Vision Transformers via Subspace Intervention
Weichen Zhou, Yawen Zou, Chunzhi Gu, Ran Dong, Haoran Xie, and Chao Zhang
L3
ECCV 2026 Oral verified in the official program; scientific claims use the conference-hosted manuscript #7101. Publisher proceedings/version of record was not verified on 2026-09-12.
Visual representations Interpretability 3D / spatial vision
visual representation learning · interpretability / attribution · 3D / spatial vision · geometric probing · subspace intervention · DINOv2 · MAE · linear readability · frozen feature diagnostics
Official artifacts: ECCV Oral program · Conference manuscript · Official implementation · arXiv manuscript
Metadata checked 2026-09-12
2026IEEE TPAMIPeer reviewed
Understanding the Effects of Projectors in Knowledge Distillation
Yudong Chen, Sen Wang, Jiajun Liu, Xuwei Xu, Frank de Hoog, Brano Kusy, and Zi Huang
L3
IEEE publisher metadata identifies TPAMI 48(8), 9078–9092, August 2026. The accessible scientific text checked here is arXiv:2310.17183v1 (26 October 2023); final journal numerical identity was not established. This extends the authors’ NeurIPS 2022 Projector Ensemble Feature Distillation work, rather than originating PEFD in 2026.
Knowledge distillation Loss / objectives Image / recognition Visual representations
knowledge distillation / model transfer · loss / objective design · image recognition · visual representation learning · PEFD · projector ensemble · feature distillation · logit distillation · optimization interference · CKA · calibration
Official artifacts: Inspected author manuscript (v1) · Author code and earlier PEFD implementation
Metadata checked 2026-10-12
2026CVPRPeer reviewed
UniM: A Unified Any-to-Any Interleaved Multimodal Benchmark
Yanlin Li et al.
L3
V&L evaluation / safety
vision–language evaluation / safety · any-to-any · interleaved multimodal · benchmark
Metadata checked 2026-07-27
2026ICLRPeer reviewed
Universal Inverse Distillation for Matching Models with Real-Data Supervision (No GANs)
Nikita Kornilov, David Li, Tikhon Mavrin, Aleksei Leonov, Nikita Gushchin, Evgeny Burnaev, Iaroslav Koshelev, Aleksandr Korotin
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Three-network training yields a one-step generator. No GAN discriminator does not mean no min–max optimization. Selected low-resolution image results include additional fine-tuning; teacher-error correction has restrictive assumptions. No cross-magnification or pathology evaluation.
Knowledge distillation Generative / representation Loss / objectives Generative vision
knowledge distillation / model transfer · generative / representation learning · loss / objective design · generative vision · RealUID · universal inverse distillation · real-data supervision · one-step generation · matching models · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview
Metadata checked 2026-09-19
2026AAAIPeer reviewed
URaG: Unified Retrieval and Generation in Multimodal LLMs for Efficient Long Document Understanding
Yongxin Shi, Jiapeng Wang, Zeyu Shan, Dezhi Peng, Zening Lin, and Lianwen Jin
L3
AAAI 2026 proceedings paper; FLOP reductions are not measured latency. Code repository states non-commercial research use.
Multimodal LLMs Document intelligence Multimodal retrieval
multimodal LLM · document intelligence · multimodal embedding / retrieval · unified retrieval generation · document understanding · page selection · late interaction · visual token pruning · LoRA
Official artifacts: Proceedings PDF · Author code and usage terms
Metadata checked 2026-10-26
2026ECCVAccepted · proceedings pending
URoPE: Universal Relative Position Embedding across Geometric Spaces
Yichen Xie, Depu Meng, Chensheng Peng, Yihan Hu, Quentin Herau, Masayoshi Tomizuka, and Wei Zhan
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #11480 checked on 2026-09-12. Publisher version of record was not independently verified.
3D / spatial vision Transformers / pretraining Graph / geometric ML
3D / spatial vision · NLP transformers / pretraining · graph / geometric learning · projective geometry · rotary position embedding · depth anchors
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026arXivPreprint
V-Rubrics: Visual Faithfulness via Rubric-Based Reinforcement Learning
Shulin Tian, Minglun Li, Yuhao Dong, Hao Ding, Jiarui Yao, Haiwen Diao, Jingkang Yang, Hongyuan Zhu, and Ziwei Liu
L2–L3
arXiv preprint first submitted 26 August 2026; not peer reviewed as of the audit date.
LLM post-training Reasoning / verification Multimodal LLMs V&L evaluation / safety
LLM post-training / alignment · LLM reasoning / verification · multimodal LLM · vision–language evaluation / safety · V-Rubrics · visual faithfulness · rubric reward · GRPO · multimodal reasoning
Official artifacts: Official project · Official code and data
Metadata checked 2026-08-30
2026arXivPreprint
VBVR-Pro: A Scalable and Verifiable Suite for Native Visual Reasoning
Junxiang Xu, Ruisi Wang, Fanyi Pu, Maijunxian Wang, Ran Ji, Tongxi Zhou, Chenyang Gu, Jing Zuo, Hongcan Xiao, Yimeng Geng, Wanqi Yin, Wei Chen, Oscar Qian, Zhengan Yan, Ziqi Huang, Haiwen Diao, Liang Pan, Bo Li, Xiangyu Fan, Dezhi Luo, Fengyuan Yu, Zehong Zhao, Qingying Gao, Tinghui Zhu, Yilan Zhang, Jingqi Tong, Pinyuan Feng, Zhengze Jiang, Letian Wang, Ziyu Guo, Renrui Zhang, Jieneng Chen, Sonia Joseph, Constantin Venhoff, Saman Motamed, Mengyue Yang, Chandra Sripada, Alan Yuille, Philip Torr, Lvmin Zhang, Vikash Kumar, Daniel Khashabi, Nikolaus Kriegeskorte, Raphaël Millière, Vincent C. Müller, Anyi Rao, Quan Wang, Ziwei Liu, Dahua Lin, Lei Yang, Hokin Deng, and Zhongang Cai
L2–L3
arXiv preprint first submitted 26 August 2026; not peer reviewed as of the audit date.
Generation / video / agents V&L evaluation / safety Multimodal LLMs Captioning / VQA
multimodal generation / agents · vision–language evaluation / safety · multimodal LLM · captioning / VQA · VBVR-Pro · native visual reasoning · verifiable rewards · procedural tasks · video generation
Official artifacts: Official project · Official code and benchmark
Metadata checked 2026-08-30
2026ACLPeer reviewed
VeriTaS: The First Dynamic Benchmark for Multimodal Automated Fact-Checking
Mark Rothermel, Marcus Kornmann, Marcus Rohrbach, and Anna Rohrbach
L2–L3
Peer-reviewed ACL 2026 long paper and Best Resource Paper award recipient.
NLP tasks / evaluation Multilingual / low resource RAG / tools / agents LLM evaluation / safety V&L evaluation / safety
NLP tasks / evaluation · multilingual / low-resource NLP · RAG / tools / agents · LLM evaluation / safety · vision–language evaluation / safety · VeriTaS · dynamic benchmark · multimodal fact-checking · data leakage · claim integrity
Official artifacts: Author manuscript · Official project and data portal · Official code
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
VesselTok: Tokenizing Vessel-like 3D Biomedical Graph Representations for Reconstruction and Generation
Chinmay Prabhakar, Bastian Wittmann, Tamaz Amiranashvili, Paul Büschl, Ezequiel De la Rosa, Julian McGinnis, Benedikt Wiestler, Bjoern Menze, and Suprosanna Shit
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #10927. Publisher proceedings/version of record was not verified on 2026-09-12.
3D / spatial vision Generative / representation Restoration / inverse problems
3D / spatial vision · generative / representation learning · VesselTok · biomedical spatial graphs · pseudo-radius occupancy · continuous graph tokenizer
Official artifacts: ECCV Spotlight program · Conference manuscript · Official implementation
Metadata checked 2026-09-12
2026CVPRPeer reviewed
VGent: Visual Grounding via Modular Design for Disentangling Reasoning and Prediction
Weitai Kang et al.
L2
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · visual grounding · modular reasoning · spatial prediction · multimodal LLM
Metadata checked 2026-07-27
2026CVPRPeer reviewed
VideoAuto-R1: Video Auto Reasoning via Thinking Once, Answering Twice
Shuming Liu et al.
L3
Generation / video / agents
multimodal generation / agents · video reasoning · reinforcement learning · adaptive inference
Metadata checked 2026-07-27
2026ACLPeer reviewed
ViLL-E: Video LLM Embeddings for Retrieval
Rohit Gupta, Jayakrishnan Unnikrishnan, Fan Fei, Sheng Liu, Son Tran, and Mubarak Shah
L2–L3
Peer-reviewed ACL 2026 long paper and Outstanding Paper award recipient.
Multimodal retrieval Generation / video / agents Multimodal LLMs
multimodal embedding / retrieval · multimodal generation / agents · multimodal LLM · ViLL-E · video-language model · adaptive embeddings · video retrieval · temporal localization
Official artifacts: Author manuscript · Official publication page
Metadata checked 2026-08-30
2026NeurIPSAccepted · proceedings pending
VIPER: An Expert-Curated Benchmark for Vision-Language Models in Veterinary Pathology
Luca L. Weishaupt, Simone de Brot, Javier Asin, Llorenç Grau-Roma, Nic G. Reitsam, Andrew H. Song, Dongmin Bang, Stefan T. Kaluziak, Long Phi Le, Jakob Nikolas Kather, Faisal Mahmood, and Guillaume Jaume
L3
Acceptance independently verified 2026-10-08 through official NeurIPS poster 139742: exact title and matching 12-author list, linked OpenReview 7986DZ9rPe. Scientific review and artifact checks remain tied to arXiv:2608.26382v1 and the 2026-08-30 audit, not a refreshed full-text/code/data inspection. VIPER evaluates rat toxicologic regions rather than whole-slide images; it is an ROI benchmark, not a WSI benchmark.
V&L evaluation / safety
pathology VLM · evaluation · robustness / domain shift / privacy · vision–language evaluation / safety · VIPER · veterinary pathology · toxicologic pathology · expert-curated VQA · visual grounding audit · domain transfer
Official artifacts: Official evaluation code and documentation · Official benchmark dataset · Official NeurIPS 2026 acceptance record
Metadata checked 2026-10-08
2026arXivPreprint
Vision as Unified Multimodal Generation
Xiaoyang Han et al.
L3
Generation / video / agents
multimodal generation / agents · SenseNova-Vision · mixed outputs · structured vision
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
Volume Transformer: Revisiting Vanilla Transformers for 3D Scene Understanding
Kadir Yilmaz, Adrian Kruse, Tristan Höfer, Daan de Geus, and Bastian Leibe
L3
Accepted to ECCV 2026; oral presentation verified in the official program. This record and review use the conference-hosted manuscript checked 2026-09-12; a separate Springer proceedings record was not verified.
3D / spatial vision Detection / segmentation Knowledge distillation
3D / spatial vision · detection / segmentation · knowledge distillation / model transfer · Volt · 3D segmentation · global attention · volumetric patch tokens · hard-label distillation · 3D RoPE
Official artifacts: Official ECCV oral presentation · Official ECCV paper record · arXiv record · Official code · Official project
Metadata checked 2026-09-12
2026ICLRPeer reviewed
WAVE: Learning Unified & Versatile Audio-Visual Embeddings with Multimodal LLM
Changli Tang, Qinfan Xiao, Ke Mei, Tianyi Wang, Fengyun Rao, Chao Zhang
L3
Official ICLR 2026 oral and full proceedings manuscript checked. Multimodal embeddings and candidate-answer ranking, not unrestricted report generation. Official paper and released training configuration checked; unequal pooling contracts, negative findings and unresolved overall-score discrepancy are documented. No WSI evaluation.
Image–text representation Multimodal retrieval Multimodal LLMs Video / 3D / embodied
image–text representation · multimodal embedding / retrieval · multimodal LLM · video / 3D / embodied vision · WAVE · multimodal embedding · audio-video retrieval · hierarchical feature fusion · prompt conditioning · ICLR 2026 oral
Official artifacts: Official proceedings record · OpenReview · Official implementation
Metadata checked 2026-09-19
2026CVPRPeer reviewed
WeDetect: Fast Open-Vocabulary Object Detection as Retrieval
Mingfei Fu et al.
L3
Detection / segmentation Grounding / open vocabulary
detection / segmentation · WeDetect · open-vocabulary detection · retrieval · grounding / open vocabulary
Metadata checked 2026-07-27
2026ECCVAccepted · proceedings pending
What CLIP Knows but Cannot Say: Recovering Negation from Frozen Intermediate Features
Chen-Yi Lu, Yueh-Shao Chen, and Somali Chaterji
L3
ECCV 2026 Oral verified in the official program (event 6084); scientific claims use the conference-hosted manuscript #10575. Publisher proceedings/version of record was not verified on 2026-09-12.
Image–text representation V&L evaluation / safety Multimodal retrieval
image–text representation · vision–language evaluation / safety · multimodal embedding / retrieval · negation · frozen feature correction · compositionality
Official artifacts: ECCV program · Conference manuscript · Author project
Metadata checked 2026-09-12
2026arXivPreprint
What Makes High-Magnification Knowledge Transferable? A Study of Cross-Resolution Distillation in Whole-Slide Imaging
Zhiyuan Yang, Jiahao Cheng, and Mahdi S. Hosseini
L3
Public arXiv v2 (2026-10-01) verified 2026-10-02. The authors state ICLR 2027 under review; no acceptance inferred. Only public source material was used. A diagnostic study of target utility, predictability and downstream benefit, not a new universally superior distillation architecture.
Knowledge distillation Evaluation / reproducibility
MIL / WSI · evaluation · foundation encoder · knowledge distillation / model transfer · evaluation / reproducibility · cross-magnification distillation · 5× to 20× · residual prediction · teacher target utility · conditional mean · WSI evaluation
Official artifacts: Reviewed manuscript revision
Metadata checked 2026-10-02
2026arXivPreprint
Whole-Slide Image Analysis under Realistic Few-Shot Annotation Protocols
Tiffanie Godelaine, Maxime Zanella, Karim El Khoury, Benoit Macq, and Christophe De Vleeschouwer
L3
arXiv v1 submitted 31 August 2026; the manuscript title page says it was submitted on 29 July 2026. Code and data instructions were released, but no peer-reviewed record was verified on 2 September 2026.
Transfer / adaptation / continual
cell / tissue modeling · pathology VLM · evaluation · SlideCRF · few-shot transduction · WSI patch classification · pathologist clicks · scribble annotation · absent classes
Official artifacts: Official implementation and data instructions
Metadata checked 2026-09-02
2026ECCVAccepted · proceedings pending
Why Far Looks Up: Probing Spatial Representation in Vision-Language Models
Cheolhong Min, Jaeyun Jung, Daeun Lee, Hyeonseong Jeon, Yu Su, Jonathan Tremblay, Chan Hee Song, and Jaesik Park
L3
ECCV 2026 Spotlight verified in the official Vision-Language Models & Visual Foundation Models session; manuscript #10791 checked on 2026-09-12. Publisher version of record was not independently verified. The project page describes an oral; the catalog follows the official Spotlight record.
V&L evaluation / safety Interpretability 3D / spatial vision
vision–language evaluation / safety · interpretability / attribution · 3D / spatial vision · SpatialTunnel · spatial shortcut · contrastive probing
Official artifacts: ECCV Spotlight program · Conference manuscript · Official project or implementation
Metadata checked 2026-09-12
2026ACLPeer reviewed
Why Multimodal In-Context Learning Lags Behind? Unveiling the Inner Mechanisms and Bottlenecks
Yu Wang and Sharon Li
L2–L3
Peer-reviewed ACL 2026 long paper.
Multimodal LLMs V&L evaluation / safety LLM interpretability
multimodal LLM · vision–language evaluation / safety · LLM interpretability / editing / unlearning · multimodal in-context learning · task mapping · mechanistic analysis · information transfer · MGI
Official artifacts: Author manuscript · Official analysis code
Metadata checked 2026-08-30
2026ACLPeer reviewed
Your Reasoning Model is Secretly a Reward Model - Optimization-Free Verification from Experience
Zhenwen Liang, Ruosen Li, Yujun Zhou, Linfeng Song, Dian Yu, Xinya Du, Haitao Mi, and Dong Yu
L2–L3
Peer-reviewed ACL 2026 long paper.
Reasoning / verification LLM evaluation / safety
LLM reasoning / verification · LLM evaluation / safety · CLUE · reasoning verification · activation delta · centroid classifier · reranking
Metadata checked 2026-08-30
2026ECCVAccepted · proceedings pending
Zero-shot Depth from Defocus
Yiming Zuo, Hongyu Wen, Venkat Subramanian, Patrick Chen, Karhan Kayan, Mario Bijelic, Felix Heide, and Jia Deng
L3
ECCV 2026 Spotlight verified in the official program, session Computational, Medical & Scientific Imaging; scientific claims use conference-hosted manuscript #3173. Publisher proceedings/version of record was not verified on 2026-09-12.
3D / spatial vision Restoration / inverse problems Vision evaluation
3D / spatial vision · vision evaluation / robustness · FOSSA · ZEDD · focus-stack attention · depth from defocus · zero-shot depth
Official artifacts: ECCV Spotlight program · Conference manuscript · Official implementation · Author project (paper/program link)
Metadata checked 2026-09-12
2025COLMPeer reviewed
2 OLMo 2 Furious (COLM’s Version)
OLMo Team
L3
The COLM 2025 paper identifies OLMo Team as the corporate author and provides the full contributor list in the paper.
LLM pretraining
LLM pretraining / scaling · OLMo 2 · open data · training recipe
Metadata checked 2026-07-28
2025Nature Biomedical EngineeringPeer reviewed
A generalizable pathology foundation model using a unified knowledge distillation pretraining framework
Ma et al.
L3
Knowledge distillation
self-supervised learning · foundation encoder · GPFM · knowledge distillation · hybrid supervision · knowledge distillation / model transfer
Metadata checked 2026-07-27
2025NaturePeer reviewed
Accurate Predictions on Small Data with a Tabular Foundation Model
Noah Hollmann et al.
L2–L3
Classical ML
classical machine learning · TabPFN · tabular foundation model · in-context learning
Metadata checked 2026-07-27
2025MICCAIPeer reviewed
AEM: Attention Entropy Maximization for Multiple Instance Learning based Whole Slide Image Classification
Yunlong Zhang, Honglin Li, Yuxuan Sun, Zhongyi Shui, Jingxiong Li, Chenglu Zhu, and Lin Yang
L3
Loss / objectives
loss / objective design · MIL / WSI · attention entropy maximization · multiple instance learning · whole-slide classification · cosine weight annealing · attention regularization
Official artifacts: Official code
Metadata checked 2026-08-10
2025ICLRPeer reviewed
Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agents
Hanrong Zhang, Jingyuan Huang, Kai Mei, Yifei Yao, Zhenting Wang, Chenlu Zhan, Hongwei Wang, and Yongfeng Zhang
L3
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · ASB · agent security · prompt injection
Official artifacts: Benchmark
Metadata checked 2026-07-28
2025CVPRPeer reviewed
Apollo: An Exploration of Video Understanding in Large Multimodal Models
Orr Zohar et al.
L2–L3
Generation / video / agents
multimodal generation / agents · video MLLM · controlled study · visual token budget
Metadata checked 2026-07-27
2025CVPRPeer reviewed
Argus: Vision-Centric Reasoning with Grounded Chain-of-Thought
Yunze Man et al.
L2
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · grounded reasoning · visual attention · chain of thought · multimodal LLM
Metadata checked 2026-07-27
2025ICLRPeer reviewed
BigCodeBench: Benchmarking Code Generation with Diverse Function Calls and Complex Instructions
Terry Yue Zhuo, Minh Chien Vu, Jenny Chim, Han Hu, Wenhao Yu, Ratnadira Widyasari, Imam Nur Bani Yusuf, Haolan Zhan, Junda He, Indraneil Paul, Simon Brunner, Chen Gong, Thong Hoang, Armel Zebaze, Xiaoheng Hong, Wen-Ding Li, Jean Kaddour, Ming Xu, Zhihan Zhang, Prateek Yadav, Naman Jain, Alex Gu, Zhoujun Cheng, Jiawei Liu, Qian Liu, Zijian Wang, Binyuan Hui, Niklas Muennighoff, David Lo, Daniel Fried, Xiaoning Du, Harm de Vries, and Leandro von Werra
L2–L3
Code / software engineering
code generation / software engineering · BigCodeBench · function-level code generation · tool use · complex instructions · execution-based evaluation
Official artifacts: Official benchmark implementation and evaluation harness · Official benchmark dataset · Author manuscript
Metadata checked 2026-08-20
2025ACLPeer reviewed
Byte Latent Transformer: Patches Scale Better Than Tokens
Artidoro Pagnoni et al.
L2–L3
ACL 2025 Outstanding Paper.
Text / sequence models Transformers / pretraining LLM pretraining
text / sequence modeling · NLP transformers / pretraining · LLM pretraining / scaling · Byte Latent Transformer · byte patches · dynamic patching · tokenization-free language modeling
Official artifacts: Official code · Official research page
Metadata checked 2026-08-10
2025WACVPeer reviewed
CEMIL: Contextual Attention Based Efficient Weakly Supervised Approach for Histopathology Image Classification
Tawsifur Rahman, Alexander S. Baras, and Rama Chellappa
L3
Peer-reviewed WACV 2025 proceedings paper, pages 4248–4257.
Knowledge distillation
knowledge distillation / model transfer · MIL / WSI · WSI engineering · CEMIL · patch selection · instructor–learner distillation · contextual attention · efficient MIL
Metadata checked 2026-08-23
2025Findings of ACLPeer reviewed
ChartQAPro: A More Diverse and Challenging Benchmark for Chart Question Answering
Ahmed Masry et al.
L3
Captioning / VQA V&L evaluation / safety Document intelligence
captioning / VQA · vision–language evaluation / safety · chart question answering · dashboard reasoning · unanswerable questions · document intelligence
Metadata checked 2026-07-28
2025ICCVPeer reviewed
CleanPose: Category-Level Object Pose Estimation via Causal Learning and Knowledge Distillation
Xiao Lin, Yun Peng, Liuyi Wang, Xianyou Zhong, Minghao Zhu, Yi Feng, Jingwei Yang, Chengju Liu, and Qijun Chen
L3
Published in the ICCV 2025 main conference proceedings, pages 5990–6000.
Knowledge distillation Causal inference 3D / spatial vision
knowledge distillation / model transfer · causal inference · 3D / spatial vision · CleanPose · category-level pose estimation · causal learning · ULIP-2 · residual adapter
Official artifacts: Official implementation
Metadata checked 2026-09-02
2025ICLRPeer reviewed
ColPali: Efficient Document Retrieval with Vision Language Models
Manuel Faysse, Hugues Sibille, Tony Wu, Bilel Omrani, Gautier Viaud, Céline Hudelot, and Pierre Colombo
L2–L3
Document intelligence Multimodal retrieval NLP tasks / evaluation
document intelligence · multimodal embedding / retrieval · NLP tasks / evaluation · ColPali · visual document retrieval · late interaction · ViDoRe
Official artifacts: Official code · Official ViDoRe models and benchmark
Metadata checked 2026-08-10
2025CVPRPeer reviewed
COSMOS: Cross-Modality Self-Distillation for Vision Language Pre-training
Sanghwan Kim et al.
L2
Image–text representation Knowledge distillation
image–text representation · cross-modal distillation · local-global views · pretraining · knowledge distillation / model transfer
Metadata checked 2026-07-27
2025CVPRPeer reviewed
Cross-modal Information Flow in Multimodal Large Language Models
Zhi Zhang et al.
L2
Multimodal LLMs
multimodal LLM · information flow · mechanistic analysis
Metadata checked 2026-07-27
2025NaturePeer reviewed
DeepSeek-R1 Incentivizes Reasoning in LLMs through Reinforcement Learning
Daya Guo et al.
L3
LLM post-training Reasoning / verification Knowledge distillation
LLM post-training / alignment · DeepSeek-R1 · RLVR · reasoning · LLM reasoning / verification · knowledge distillation / model transfer
Metadata checked 2026-07-27
2025ICMLPeer reviewed
Distillation Scaling Laws
Dan Busbridge, Amitis Shidani, Floris Weers, Jason Ramapuram, Etai Littwin, and Russ Webb
L2–L3
Knowledge distillation LLM pretraining
knowledge distillation / model transfer · LLM pretraining / scaling · distillation scaling laws · compute allocation · teacher–student scaling
Metadata checked 2026-08-10
2025MICCAIPeer reviewed
Distilling Foundation Models for Robust and Efficient Models in Digital Pathology
Alexandre Filiot, Nicolas Dop, Oussama Tchita, Auriane Riou, Rémy Dubois, Thomas Peeters, Daria Valter, Marin Scalbert, Charlie Saillard, Geneviève Robin, and Antoine Olivier
L3
Peer-reviewed MICCAI 2025 proceedings paper.
Knowledge distillation
knowledge distillation / model transfer · foundation encoder · evaluation · robustness / domain shift / privacy · H0-mini · foundation-model distillation · compact encoder · scanner robustness · stain robustness
Official artifacts: Official model · Official robustness benchmark
Metadata checked 2026-08-23
2025CVPRPeer reviewed
F-LMM: Grounding Frozen Large Multimodal Models
Size Wu et al.
L3
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · multimodal LLM · frozen MLLM · attention grounding · segmentation
Metadata checked 2026-07-27
2025CVPRPeer reviewed
Fast and Accurate Gigapixel Pathological Image Classification with Hierarchical Distillation Multi-Instance Learning
Jiuyang Dong et al.
L3
Knowledge distillation
MIL / WSI · hierarchical distillation · multi-resolution · efficiency · knowledge distillation / model transfer
Official artifacts: Code
Metadata checked 2026-07-27
2025arXivTechnical report
Gemma 3 Technical Report
Gemma Team et al.
L2
Multimodal LLMs
multimodal LLM · long context · distillation
Metadata checked 2026-07-27
2025Scientific ReportsPeer reviewed
Human Visual Attention-Inspired Knowledge Distillation Underlying Interpretable Computational Pathology
Muzhou Yu, Zihan Zhong, Xingang Zhou, Yuekun Wang, Tingyu Liang, Jiamin Chen, Hongmin Huang, Junhan Zhou, Dachun Zhao, Bo Lei, Yu Wang, Wenbin Ma, and Kaisheng Ma
L3
Peer-reviewed open-access journal article 15:42124. Low- and high-magnification language motivates the design, but HVisKD transfers sample- and region-relation features between teacher and student given the same pathology patches; it is not cross-magnification distillation.
Knowledge distillation Interpretability
knowledge distillation / model transfer · interpretability / attribution · cell / tissue modeling · WSI engineering · evaluation · HVisKD · relation distillation · interpretable distillation · GBM tissue mapping · lightweight CNN
Official artifacts: Open-access article · Official code
Metadata checked 2026-09-02
2025arXivTechnical report
InternVideo2.5: Empowering Video MLLMs with Long and Rich Context Modeling
Yi Wang et al.
L2
Generation / video / agents
multimodal generation / agents · long video · memory · token compression
Metadata checked 2026-07-27
2025arXivPreprint
InternVL3: Exploring Advanced Training and Test-Time Recipes for Open-Source Multimodal Models
Jinguo Zhu et al.
L3
Multimodal LLMs
multimodal LLM · native multimodal pretraining · preference optimization · test-time scaling
Metadata checked 2026-07-27
2025ICMLPeer reviewed
Interpreting CLIP with Hierarchical Sparse Autoencoders
Vladimir Zaigrajew, Hubert Baniecki, and Przemyslaw Biecek
L3
Zaigrajew et al. multi-TopK CLIP method, distinct from Bussmann et al. fixed-prefix Matryoshka SAE. Author repository discloses a post-acceptance text-standardization correction; review distinguishes proceedings measurements from revised code.
Image–text representation Visual representations Interpretability
image–text representation · visual representation learning · interpretability / attribution · SAE · Matryoshka SAE for CLIP · sparse autoencoders · multiple TopK budgets · concept discovery · progressive recovery
Official artifacts: Proceedings PDF · Author code and correction notice · Original ICML implementation · arXiv
Metadata checked 2026-10-09
2025arXivPreprint
Janus-Pro: Unified Multimodal Understanding and Generation with Data and Model Scaling
Xiaokang Chen et al.
L2
Generation / video / agents
multimodal generation / agents · unified model · decoupled pathways · image generation
Metadata checked 2026-07-27
2025CVPRPeer reviewed
JanusFlow: Harmonizing Autoregression and Rectified Flow for Unified Multimodal Understanding and Generation
Yiyang Ma et al.
L2–L3
Generation / video / agents
multimodal generation / agents · rectified flow · multimodal unification · autoregression
Metadata checked 2026-07-28
2025ICLRPeer reviewed
KAN: Kolmogorov–Arnold Networks
Ziming Liu et al.
L3
Neural networks
neural networks · KAN · learned edge functions · function approximation
Metadata checked 2026-07-27
2025NeurIPSPeer reviewed
Large Language Diffusion Models
Shen Nie et al.
L2–L3
Generative / representation Transformers / pretraining LLM pretraining
generative / representation learning · NLP transformers / pretraining · LLM pretraining / scaling · LLaDA · masked diffusion language model · non-autoregressive language modeling
Official artifacts: Official code and checkpoints
Metadata checked 2026-08-10
2025ICMLPeer reviewed
Learning Multi-Level Features with Matryoshka Sparse Autoencoders
Bart Bussmann, Noa Nabeshima, Adam Karvonen, and Neel Nanda
L3
Interpretability
interpretability / attribution · SAE · Matryoshka SAE · BatchTopK · nested dictionaries · feature absorption · hierarchical concepts
Official artifacts: Author code · Toy-model code · Author manuscript
Metadata checked 2026-10-09
2025NeurIPSPeer reviewed
Learning Task-Agnostic Representations through Multi-Teacher Distillation
Philippe Formont, Maxime Darrin, Banafsheh Karimian, Eric Granger, Jackie CK Cheung, Ismail Ben Ayed, Mohammadhadi Shateri, and Pablo Piantanida
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · multi-teacher distillation · task-agnostic representation · mutual information · Gaussian kernels · embedding compression
Official artifacts: Official text-model code · Official vision-model code · Official molecular-model code
Metadata checked 2026-08-10
2025ICMLPeer reviewed
Learning to (Learn at Test Time): RNNs with Expressive Hidden States
Yu Sun, Xinhao Li, Karan Dalal, Jiarui Xu, Arjun Vikram, Genghan Zhang, Yann Dubois, Xinlei Chen, Xiaolong Wang, Sanmi Koyejo, Tatsunori Hashimoto, and Carlos Guestrin
L2–L3
Neural networks LLM pretraining LLM systems
neural networks · LLM pretraining / scaling · LLM training / serving systems · test-time training layers · online self-supervised learning · expressive hidden state · long-context sequence modeling
Official artifacts: Official code
Metadata checked 2026-08-10
2025arXivPreprint
LeJEPA: Provable and Scalable Self-Supervised Learning Without the Heuristics
Randall Balestriero and Yann LeCun
L3
Preprint revised through 2026-05-25; no peer-reviewed version of record was located by 2026-08-30.
Generative / representation Visual representations
generative / representation learning · visual representation learning · LeJEPA · joint embedding · SIGReg · self-supervised learning · collapse prevention
Official artifacts: Official implementation
Metadata checked 2026-08-30
2025ICLRPeer reviewed
LiveBench: A Challenging, Contamination-Limited LLM Benchmark
Colin White et al.
L2–L3
LLM evaluation / safety
LLM evaluation / safety · LiveBench · dynamic benchmark · contamination
Official artifacts: Benchmark · Leaderboard
Metadata checked 2026-07-28
2025ICLRPeer reviewed
LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code
Naman Jain, King Han, Alex Gu, Wen-Ding Li, Fanjia Yan, Tianjun Zhang, Sida Wang, Armando Solar-Lezama, Koushik Sen, and Ion Stoica
L2–L3
Code / software engineering
code generation / software engineering · LiveCodeBench · dynamic benchmark · contamination control · code generation · self-repair
Official artifacts: Official benchmark implementation
Metadata checked 2026-08-20
2025ICLRPeer reviewed
LLaVA-MoD: Making LLaVA Tiny via MoE-Knowledge Distillation
Fangxun Shu, Yue Liao, Lei Zhang, Le Zhuo, Chenning Xu, Guanghao Zhang, Haonan Shi, Long Chen, Tao Zhong, Wanggui He, Siming Fu, Haoyuan Li, Zhelun Yu, Si Liu, Hongsheng Li, and Hao Jiang
L2–L3
Peer-reviewed ICLR 2025 conference paper.
Knowledge distillation Multimodal LLMs LLM systems LLM post-training
knowledge distillation / model transfer · multimodal LLM · LLM training / serving systems · LLM post-training / alignment · LLaVA-MoD · multimodal distillation · sparse mixture of experts · preference distillation · small MLLM
Official artifacts: Official code
Metadata checked 2026-08-23
2025TMLRPeer reviewed
LLaVA-OneVision: Easy Visual Task Transfer
Bo Li et al.
L2–L3
First released as a 2024 preprint; subsequently published in TMLR.
Multimodal LLMs
multimodal LLM · LLaVA-OneVision · multi-image · video understanding
Official artifacts: Official code
Metadata checked 2026-08-08
2025CVPRPeer reviewed
Magma: A Foundation Model for Multimodal AI Agents
Jianwei Yang et al.
L3
Generation / video / agents
multimodal generation / agents · multimodal agent · GUI · robotics · action grounding
Metadata checked 2026-07-27
2025arXivPreprint
Magnification-Aware Distillation (MAD): A Self-Supervised Framework for Unified Representation Learning in Gigapixel Whole-Slide Images
Mahmut S. Gokmen, Mitchell A. Klusty, Peter T. Nelson, Allison M. Neltner, Sen-Ching Samson Cheung, Thomas M. Pearce, David A. Gutman, Brittany N. Dugger, Devavrat S. Bisht, Margaret E. Flanagan, and V. K. Cody Bumgardner
L3
Preprint; arXiv comments report submission to the AMIA 2026 Informatics Summit, not acceptance.
Knowledge distillation
knowledge distillation / model transfer · foundation encoder · self-supervised learning · WSI engineering · MAD · magnification-aware distillation · resolution-invariant embeddings · neuropathology
Official artifacts: Official model
Metadata checked 2026-08-23
2025CVPRPeer reviewed
MambaVision: A Hybrid Mamba-Transformer Vision Backbone
Ali Hatamizadeh and Jan Kautz
L3
Visual representations
visual representation learning · MambaVision · state-space model · vision transformer
Metadata checked 2026-07-27
2025NaturePeer reviewed
Mastering Diverse Control Tasks through World Models
Danijar Hafner et al.
L3
AI / reinforcement learning
AI / reinforcement learning · DreamerV3 · world model · model-based reinforcement learning
Metadata checked 2026-07-27
2025NeurIPSPeer reviewed
Mean Flows for One-step Generative Modeling
Zhengyang Geng et al.
L3
Generative / representation Generative vision
generative / representation learning · MeanFlow · one-step generation · flow matching · generative vision
Metadata checked 2026-07-27
2025ICLRPeer reviewed
MiniPLM: Knowledge Distillation for Pre-Training Language Models
Yuxian Gu, Hao Zhou, Fandong Meng, Jie Zhou, and Minlie Huang
L2–L3
Peer-reviewed ICLR 2025 conference paper.
Knowledge distillation Transformers / pretraining LLM pretraining LLM systems
knowledge distillation / model transfer · NLP transformers / pretraining · LLM pretraining / scaling · LLM training / serving systems · MiniPLM · pretraining distillation · offline teacher scoring · data distribution refinement · cross-tokenizer transfer
Official artifacts: Official code, data, and models
Metadata checked 2026-08-23
2025ICLRPeer reviewed
MMTEB: Massive Multilingual Text Embedding Benchmark
Kenneth Enevoldsen et al.
L2
Review uses the proceedings PDF and Table 2 benchmark definitions. The landing abstract states 1,000+ languages while the PDF abstract states 250+; Section 5 attributes the broader count to bitext mining. Task-count inconsistencies in prose are retained as caveats, not silently combined.
NLP tasks / evaluation Multilingual / low resource RAG / tools / agents
NLP tasks / evaluation · multilingual / low-resource NLP · RAG / tools / agents · multilingual embeddings · retrieval evaluation · task selection · language coverage · benchmark efficiency
Official artifacts: Proceedings PDF · Benchmark implementation · Versioned evaluation results · arXiv
Metadata checked 2026-10-12
2025CVPRPeer reviewed
Molmo and PixMo: Open Weights and Open Data for State-of-the-Art Vision-Language Models
Matt Deitke et al.
L2–L3
First released as an arXiv preprint in 2024; the version of record was published at CVPR 2025.
Captioning / VQA Multimodal LLMs Grounding / open vocabulary V&L evaluation / safety
captioning / VQA · multimodal LLM · grounding / open vocabulary · vision–language evaluation / safety · open data · dense captions · VQA · pointing
Official artifacts: Project and artifacts
Metadata checked 2026-07-28
2025CVPRPeer reviewed
MonoTAKD: Teaching Assistant Knowledge Distillation for Monocular 3D Object Detection
Hou-I Liu, Christine Wu, Jen-Hao Cheng, Wenhao Chai, Shian-Yun Wang, Gaowen Liu, Hugo Latapie, Jhih-Ciang Wu, Jenq-Neng Hwang, Hong-Han Shuai, and Wen-Huang Cheng
L3
Published in the CVPR 2025 main conference proceedings, pages 22266–22275.
Knowledge distillation 3D / spatial vision Detection / segmentation
knowledge distillation / model transfer · 3D / spatial vision · detection / segmentation · MonoTAKD · monocular 3D detection · cross-modal residual distillation · LiDAR teacher · BEV
Official artifacts: Official implementation and models
Metadata checked 2026-09-02
2025MICCAIPeer reviewed
Multi-modal Knowledge Decomposition based Online Distillation for Biomarker Prediction in Breast Cancer Histopathology
Qibin Zhang, Xinyu Hao, Qiao Chen, Rui Xu, Fengyu Cong, Cheng Lu, and Hongming Xu
L3
Knowledge distillation
knowledge distillation / model transfer · spatial / molecular · MIL / WSI · whole-slide image · genomics · biomarker prediction · online distillation · knowledge decomposition
Official artifacts: Official code
Metadata checked 2026-08-10
2025CVPRPeer reviewed
Multimodal Autoregressive Pre-training of Large Vision Encoders
Enrico Fini et al.
L3
Visual representations Image–text representation
visual representation learning · AIMv2 · autoregressive pretraining · vision encoder · image–text representation
Metadata checked 2026-07-27
2025Moonshot AI technical reportTechnical report
Muon is Scalable for LLM Training
Jingyuan Liu et al.
L3
The manuscript self-identifies as a technical report; no peer-reviewed venue was verified through 2026-07-28.
Neural networks
neural networks · Muon · matrix optimization · scaling
Metadata checked 2026-07-28
2025ACLPeer reviewed
Native Sparse Attention: Hardware-Aligned and Natively Trainable Sparse Attention
Jingyang Yuan et al.
L2–L3
ACL 2025 Best Paper.
Transformers / pretraining LLM pretraining LLM systems
NLP transformers / pretraining · LLM pretraining / scaling · LLM training / serving systems · Native Sparse Attention · hardware-aligned sparse attention · long context
Metadata checked 2026-08-10
2025CVPRPeer reviewed
OmniDocBench: Benchmarking Diverse PDF Document Parsing with Comprehensive Annotations
Linke Ouyang et al.
L2
Review pins the CVPR 2025 paper and supplement (981 pages). The live repository has later data, matching and scoring revisions; its current leaderboard is not the paper benchmark.
Document intelligence V&L evaluation / safety
document intelligence · vision–language evaluation / safety · document parsing · OCR · reading order · table recognition · formula recognition · benchmark versioning
Official artifacts: Proceedings PDF · Supplement · Author code and version history
Metadata checked 2026-10-12
2025WACVPeer reviewed
On Explaining Knowledge Distillation: Measuring and Visualising the Knowledge Transfer Process
Gereziher Adhane, Mohammad Mahdi Dehshibi, Dennis Vetter, David Masip, and Gemma Roig
L3
Published in the WACV 2025 proceedings, pages 3467–3476.
Knowledge distillation Interpretability Evaluation / reproducibility Image / recognition
knowledge distillation / model transfer · interpretability / attribution · evaluation / reproducibility · image recognition · UniCAM · knowledge-distillation explainability · residual features · feature similarity score · relevance score
Official artifacts: Official implementation
Metadata checked 2026-09-02
2025ICMLPeer reviewed
On Temperature Scaling and Conformal Prediction of Deep Classifiers
Lahav Dabah and Tom Tirer
L3
Math / statistics Evaluation / reproducibility Vision evaluation
mathematics / statistics · evaluation / reproducibility · vision evaluation / robustness · temperature scaling · conformal prediction · APS · RAPS · conditional coverage · prediction set size · calibration
Official artifacts: Final proceedings PDF · Author implementation · Author manuscript
Metadata checked 2026-10-12
2025arXivPreprint
Open-Qwen2VL: Compute-Efficient Pre-Training of Fully-Open Multimodal LLMs on Academic Resources
Weizhi Wang et al.
L2
Multimodal LLMs
multimodal LLM · open pretraining · compute efficiency · reproduction
Metadata checked 2026-07-27
2025CoRL 2024 / PMLRPeer reviewed
OpenVLA: An Open-Source Vision-Language-Action Model
Moo Jin Kim et al.
L2–L3
The conference was held in November 2024; the PMLR version of record was published in 2025.
Generation / video / agents Video / 3D / embodied
multimodal generation / agents · vision-language-action · robot manipulation · open weights · video / 3D / embodied vision
Metadata checked 2026-07-28
2025IEEE AccessPeer reviewed
PatHME: Hierarchical Multi-Expert Knowledge Distillation for Whole Slide Image Analysis
Jing Wei Tan, Gangsu Kim, and Won-Ki Jeong
L3
Peer-reviewed journal article 13:163246–163257.
Knowledge distillation
knowledge distillation / model transfer · MIL / WSI · foundation encoder · WSI engineering · PatHME · multi-expert distillation · prompt-guided multiscale transfer · hierarchical attention · pathology foundation models
Official artifacts: Official code
Metadata checked 2026-08-23
2025NeurIPSPeer reviewed
Perception Encoder: The Best Visual Embeddings Are Not at the Output of the Network
Daniel Bolya et al.
L3
Visual representations
visual representation learning · Perception Encoder · vision-language alignment · intermediate features
Metadata checked 2026-07-27
2025CVPRPeer reviewed
PhD: A ChatGPT-Prompted Visual Hallucination Evaluation Dataset
Jiazhen Liu et al.
L3
V&L evaluation / safety
vision–language evaluation / safety · PhD benchmark · visual hallucination · counter-common-sense
Metadata checked 2026-07-27
2025MICCAIPeer reviewed
Prototype-Based Multiple Instance Learning for Gigapixel Whole Slide Image Classification
Susu Sun, Dominique van Midden, Geert Litjens, and Christian F. Baumgartner
L3
Official MICCAI 2025 record and 11-page accepted manuscript checked. PANDA is evaluated as binary tumor classification in this study, not the original six-class ISUP grading task. Artifact concepts are masked before MIL retraining; this is not a universal AUC improvement.
Interpretability
MIL / WSI · evaluation · interpretability / attribution · SAE · ProtoMIL · sparse autoencoders · concept intervention · additive MIL · CONCH
Official artifacts: Official MICCAI open-access record · Official accepted manuscript · Author implementation · arXiv manuscript
Metadata checked 2026-10-09
2025arXivTechnical report
Qwen2.5-VL Technical Report
Shuai Bai et al.
L2–L3
Multimodal LLMs Captioning / VQA Grounding / open vocabulary Generation / video / agents
multimodal LLM · captioning / VQA · documents · grounding · long video · visual agent · grounding / open vocabulary · multimodal generation / agents
Metadata checked 2026-07-27
2025arXivTechnical report
Qwen3 Technical Report
Qwen Team
L3
LLM pretraining LLM post-training
LLM pretraining / scaling · LLM post-training / alignment · Qwen3 · mixture of experts · thinking budget
Metadata checked 2026-07-27
2025CVPRPeer reviewed
RADIOv2.5: Improved Baselines for Agglomerative Vision Foundation Models
Greg Heinrich et al.
L3
Visual representations Knowledge distillation
visual representation learning · RADIOv2.5 · multi-teacher distillation · vision foundation model · knowledge distillation / model transfer
Metadata checked 2026-07-27
2025NeurIPSPeer reviewed
RiOSWorld: Benchmarking the Risk of Multimodal Computer-Use Agents
Jingyi Yang et al.
L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · computer-use agents · risk evaluation · interactive environment
Metadata checked 2026-07-28
2025ICMLPeer reviewed
SAEBench: A Comprehensive Benchmark for Sparse Autoencoders in Language Model Interpretability
Adam Karvonen, Can Rager, Johnny Lin, Curt Tigges, Joseph Isaac Bloom, David Chanin, Yeu-Tong Lau, Eoin Farrell, Callum Stuart Mcdougall, Kola Ayonrinde, Demian Till, Matthew Wearden, Arthur Conmy, Samuel Marks, and Neel Nanda
L3
Interpretability
interpretability / attribution · SAE · SAEBench · sparse autoencoder evaluation · feature absorption · concept isolation · spurious correlation removal
Official artifacts: Official evaluation code · Author manuscript · Interactive evaluations
Metadata checked 2026-10-09
2025ICLRPeer reviewed
SAM 2: Segment Anything in Images and Videos
Nikhila Ravi et al.
L3
Detection / segmentation Video / 3D / embodied Grounding / open vocabulary
detection / segmentation · video / 3D / embodied vision · SAM 2 · promptable segmentation · video segmentation · grounding / open vocabulary
Metadata checked 2026-07-27
2025ICMLPeer reviewed
Scalable Generation of Spatial Transcriptomics from Histology Images via Whole-Slide Flow Matching
Tinglin Huang, Tianyu Liu, Mehrtash Babadi, Wengong Jin, and Rex Ying
L3
Generative / representation Generative vision
spatial / molecular · generative / virtual pathology · STFlow · whole-slide flow matching · spatial transcriptomics · local spatial attention · generative / representation learning · generative vision
Official artifacts: Official code
Metadata checked 2026-08-10
2025ICLRPeer reviewed
Scaling and evaluating sparse autoencoders
Leo Gao, Tom Dupré la Tour, Henk Tillman, Gabriel Goh, Rajan Troll, Alec Radford, Ilya Sutskever, Jan Leike, and Jeffrey Wu
L3
Interpretability
interpretability / attribution · SAE · TopK SAE · sparse autoencoders · activation shrinkage · AuxK · mechanistic interpretability · SAE evaluation
Official artifacts: Official code · Author manuscript (2024)
Metadata checked 2026-10-09
2025ICCVPeer reviewed
Scaling Language-Free Visual Representation Learning
David Fan et al.
L3
Visual representations
visual representation learning · visual SSL · scaling · language-free pretraining
Metadata checked 2026-07-27
2025ICLRPeer reviewed
Scaling LLM Test-Time Compute Optimally Can Be More Effective than Scaling Model Parameters
Charlie Snell, Jaehoon Lee, Kelvin Xu, and Aviral Kumar
L3
LLM post-training Reasoning / verification
LLM post-training / alignment · test-time compute · verifier · compute allocation · LLM reasoning / verification
Metadata checked 2026-07-28
2025ICLRPeer reviewed
Show-o: One Single Transformer to Unify Multimodal Understanding and Generation
Jinheng Xie et al.
L2
Generation / video / agents
multimodal generation / agents · unified model · discrete diffusion · autoregression
Metadata checked 2026-07-27
2025CVPRPeer reviewed
ShowUI: One Vision-Language-Action Model for GUI Visual Agent
Kevin Qinghong Lin et al.
L3
Generation / video / agents
multimodal generation / agents · GUI agent · visual token selection · action streaming
Metadata checked 2026-07-27
2025arXivPreprint
SigLIP 2: Multilingual Vision-Language Encoders with Improved Semantic Understanding, Localization, and Dense Features
Michael Tschannen et al.
L2–L3
Image–text representation Grounding / open vocabulary Multimodal retrieval
image–text representation · grounding / open vocabulary · sigmoid alignment · multilingual · dense features · data curation · multimodal embedding / retrieval
Official artifacts: Models
Metadata checked 2026-07-27
2025ICLRPeer reviewed
Sparse Autoencoders Do Not Find Canonical Units of Analysis
Patrick Leask, Bart Bussmann, Michael Pearce, Joseph Bloom, Curt Tigges, Noura Al Moubayed, Lee Sharkey, and Neel Nanda
L3
Interpretability
interpretability / attribution · SAE · BatchTopK · meta-SAEs · SAE stitching · feature composition · dictionary size · interpretability evaluation
Official artifacts: Author manuscript · BatchTopK component implementation · Author meta-SAE dashboard
Metadata checked 2026-10-09
2025NeurIPSPeer reviewed
Sparse Autoencoders Learn Monosemantic Features in Vision-Language Models
Mateusz Pach, Shyamgopal Karthik, Quentin Bouniot, Serge Belongie, and Zeynep Akata
L3
NeurIPS 2025 proceedings volume 38; the repository BibTeX uses publication year 2026. Catalog year follows the conference edition.
Image–text representation V&L evaluation / safety Multimodal LLMs Interpretability
image–text representation · vision–language evaluation / safety · multimodal LLM · interpretability / attribution · SAE · sparse autoencoders · monosemanticity · human evaluation · visual steering · SAE benchmark
Official artifacts: Proceedings PDF · Code and human-comparison benchmark · arXiv
Metadata checked 2026-10-09
2025ICLRPeer reviewed
Sparse autoencoders reveal selective remapping of visual concepts during adaptation
Hyesu Lim, Jinho Choi, Jaegul Choo, and Steffen Schneider
L3
Visual representations Image–text representation Transfer / adaptation / continual Interpretability
visual representation learning · image–text representation · interpretability / attribution · SAE · PatchSAE · sparse autoencoders · spatial attribution · prompt adaptation · concept remapping
Official artifacts: Proceedings PDF · Code and demo
Metadata checked 2026-10-09
2025CVPRPeer reviewed
T2ISafety: Benchmark for Assessing Fairness, Toxicity, and Privacy in Image Generation
Lijun Li et al.
L3
V&L evaluation / safety Generation / video / agents
vision–language evaluation / safety · multimodal generation / agents · text-to-image safety · fairness · toxicity · privacy
Metadata checked 2026-07-28
2025ICLRPeer reviewed
TAID: Temporally Adaptive Interpolated Distillation for Efficient Knowledge Transfer in Language Models
Makoto Shing, Kou Misaki, Han Bao, Sho Yokoi, and Takuya Akiba
L2–L3
Knowledge distillation Generation / MT evaluation LLM post-training LLM systems
knowledge distillation / model transfer · generation / translation evaluation · LLM post-training / alignment · LLM training / serving systems · adaptive interpolation · language-model distillation · capacity gap · forward KL
Official artifacts: Official code
Metadata checked 2026-08-10
2025ICLRPeer reviewed
Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model
Chunting Zhou et al.
L3
Generation / video / agents
multimodal generation / agents · unified multimodal · diffusion · next-token prediction
Metadata checked 2026-07-27
2025ICMLPeer reviewed
Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment
Harrish Thasarathan, Julian Forsyth, Thomas Fel, Matthew Kowal, and Konstantinos G. Derpanis
L3
Visual representations Image–text representation Interpretability
visual representation learning · image–text representation · interpretability / attribution · SAE · universal sparse autoencoders · cross-model concepts · representation alignment · shared dictionary
Official artifacts: Proceedings PDF · Code · Project
Metadata checked 2026-10-09
2025MICCAIPeer reviewed
Unsupervised Anomaly Detection on Preclinical Liver H&E Whole Slide Images Using Graph-Based Feature Distillation
Lin Li, Lillie Shelton, Thomas Forest, Kyathanahalli Janardhan, Tiffany Jenkins, Michael J. Napolitano, Roujia Wang, David Leigh, Tosha Shah, Grady Earl Carlson, Rajath Soans, and Antong Chen
L3
Peer-reviewed MICCAI 2025 proceedings paper; evaluated on preclinical rat-liver H&E WSIs. The author project page requests email contact for code access and showed code as coming soon when checked on 2026-08-23.
Knowledge distillation
knowledge distillation / model transfer · WSI engineering · evaluation · GraphTox · graph feature distillation · unsupervised anomaly detection · preclinical toxicology · resolution-aware encoder
Official artifacts: Author project and code-access page
Metadata checked 2026-08-23
2025arXivTechnical report
V-JEPA 2: Self-Supervised Video Models Enable Understanding, Prediction and Planning
Yann LeCun et al.
L3
Video / 3D / embodied
video / 3D / embodied vision · V-JEPA 2 · world model · planning
Metadata checked 2026-07-27
2025CVPRPeer reviewed
VGGT: Visual Geometry Grounded Transformer
Jianyuan Wang et al.
L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · VGGT · feed-forward geometry · 3D reconstruction · 3D / spatial vision
Metadata checked 2026-07-27
2025CVPRPeer reviewed
Video-MME: The First-Ever Comprehensive Evaluation Benchmark of Multi-modal LLMs in Video Analysis
Chaoyou Fu et al.
L2–L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · video benchmark · duration · subtitles
Metadata checked 2026-07-27
2025CVPRPeer reviewed
Video-XL: Extra-Long Vision Language Model for Hour-Scale Video Understanding
Yan Shu et al.
L2
Generation / video / agents
multimodal generation / agents · long video · visual summarization token · compression
Metadata checked 2026-07-27
2025CVPRPeer reviewed
VidHalluc: Evaluating Temporal Hallucinations in Multimodal Large Language Models for Video Understanding
Chaoyu Li, Eun Woo Im, and Pooyan Fazli
L3
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · video hallucination · temporal order · scene transition
Metadata checked 2026-07-27
2025CVPRPeer reviewed
VL-RewardBench: A Challenging Benchmark for Vision-Language Generative Reward Models
Lei Li et al.
L3
V&L evaluation / safety
vision–language evaluation / safety · reward model · hallucination detection · visual reasoning
Metadata checked 2026-07-27
2025ICLRPeer reviewed
VLM2Vec: Training Vision-Language Models for Massive Multimodal Embedding Tasks
Ting Jiang et al.
L2–L3
Multimodal retrieval Image–text representation V&L evaluation / safety
multimodal embedding / retrieval · image–text representation · vision–language evaluation / safety · VLM2Vec · MMEB · instruction-conditioned embedding
Official artifacts: Official code · Project page
Metadata checked 2026-08-08
2025NeurIPSPeer reviewed
Why Knowledge Distillation Works in Generative Models: A Minimal Working Explanation
Sungmin Cha and Kyunghyun Cho
L2–L3
Knowledge distillation Generative / representation Generation / MT evaluation
knowledge distillation / model transfer · generative / representation learning · generation / translation evaluation · generative distillation · precision-recall trade-off · teacher entropy · Gaussian mixtures · language models
Official artifacts: Official code
Metadata checked 2026-08-10
2025ICMLPeer reviewed
μnit Scaling: Simple and Scalable FP8 LLM Training
Saaketh Narayan, Abhay Gupta, Mansheej Paul, and Davis Blalock
L3
Neural networks
neural networks · μnit Scaling · FP8 · training systems
Metadata checked 2026-07-27
2024IEEE Transactions on Medical ImagingPeer reviewed
A Graph-Based Multi-Scale Approach With Knowledge Distillation for WSI Classification
Gianpaolo Bontempo, Federico Bolelli, Angelo Porrello, Simone Calderara, and Elisa Ficarra
L3
Peer-reviewed journal article 43(4):1412–1421; journal extension of DAS-MIL.
Knowledge distillation
knowledge distillation / model transfer · MIL / WSI · WSI engineering · DAS-MIL · multiscale graph · self-distillation · spatial correlation · whole-slide classification
Official artifacts: Official code
Metadata checked 2026-08-23
2024ICLRPeer reviewed
AgentBench: Evaluating LLMs as Agents
Xiao Liu et al.
L2–L3
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · AgentBench · agent evaluation · interactive environment
Official artifacts: Benchmark
Metadata checked 2026-07-28
2024NeurIPSPeer reviewed
Are We on the Right Way for Evaluating Large Vision-Language Models?
Lin Chen et al.
L3
V&L evaluation / safety
vision–language evaluation / safety · MMStar · visual dependence · benchmark leakage
Metadata checked 2026-07-27
2024NeurIPSPeer reviewed
Autoregressive Image Generation without Vector Quantization
Tianhong Li et al.
L3
Generative vision
generative vision · MAR · continuous tokens · autoregressive generation
Metadata checked 2026-07-27
2024MLSysPeer reviewed
AWQ: Activation-Aware Weight Quantization for On-Device LLM Compression and Acceleration
Ji Lin, Jiaming Tang, Haotian Tang, Shang Yang, Wei-Ming Chen, Wei-Chen Wang, Guangxuan Xiao, Xingyu Dang, Chuang Gan, and Song Han
L2–L3
LLM systems LLM evaluation / safety
LLM training / serving systems · LLM evaluation / safety · AWQ · weight-only quantization · activation awareness · on-device inference
Official artifacts: Official project page · Official implementation and model zoo
Metadata checked 2026-08-10
2024ECCVPeer reviewed
BLINK: Multimodal Large Language Models Can See but Not Perceive
Xingyu Fu et al.
L2
V&L evaluation / safety
vision–language evaluation / safety · visual perception · diagnostic benchmark · multi-image reasoning
Metadata checked 2026-07-28
2024NeurIPSPeer reviewed
Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs
Shengbang Tong et al.
L2–L3
Multimodal LLMs
multimodal LLM · vision-centric MLLM · spatial vision aggregator · open recipe
Metadata checked 2026-07-28
2024arXivPreprint
Chameleon: Mixed-Modal Early-Fusion Foundation Models
Chameleon Team
L2–L3
Generation / video / agents
multimodal generation / agents · early fusion · interleaved generation · visual tokens
Metadata checked 2026-07-27
2024NeurIPSPeer reviewed
CogVLM: Visual Expert for Pretrained Language Models
Weihan Wang et al.
L2
First released as an arXiv preprint in 2023; the version of record was published at NeurIPS 2024.
Multimodal LLMs
multimodal LLM · visual expert · fusion
Metadata checked 2026-07-28
2024NeurIPSPeer reviewed
Compact Language Models via Pruning and Knowledge Distillation
Saurav Muralidharan, Sharath Turuvekere Sreenivas, Raviraj Joshi, Marcin Chochowski, Mostofa Patwary, Mohammad Shoeybi, Bryan Catanzaro, Jan Kautz, and Pavlo Molchanov
L3
Published in the NeurIPS 2024 main conference proceedings.
Knowledge distillation LLM systems
knowledge distillation / model transfer · LLM training / serving systems · Minitron · structured pruning · knowledge distillation · compact language model · model family
Official artifacts: Official implementation · Official model collection
Metadata checked 2026-08-30
2024NeurIPS Datasets and BenchmarksPeer reviewed
CRAG: A Comprehensive RAG Benchmark
Weijia Shi et al.
L3
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · CRAG · retrieval-augmented generation · benchmark
Official artifacts: Benchmark
Metadata checked 2026-07-28
2024ICLRPeer reviewed
Data Filtering Networks
Alex Fang et al.
L2
Image–text representation Multimodal retrieval
image–text representation · data curation · filtering network · image-text data · multimodal embedding / retrieval
Metadata checked 2026-07-27
2024arXivTechnical report
DeepSeek-V3 Technical Report
DeepSeek-AI
L3
LLM pretraining
LLM pretraining / scaling · DeepSeek-V3 · mixture of experts · multi-token prediction
Metadata checked 2026-07-27
2024arXivPreprint
DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
Zhihong Shao, Peiyi Wang, Qihao Zhu, Runxin Xu, Junxiao Song, Xiao Bi, Haowei Zhang, Mingchuan Zhang, Y. K. Li, Y. Wu, and Daya Guo
L2–L3
Reasoning / verification LLM post-training
LLM reasoning / verification · LLM post-training / alignment · DeepSeekMath · mathematical reasoning · GRPO · continued pretraining
Official artifacts: Official models and implementation
Metadata checked 2026-08-10
2024ICLRPeer reviewed
Demystifying CLIP Data
Hu Xu et al.
L2–L3
Image–text representation V&L evaluation / safety
image–text representation · vision–language evaluation / safety · MetaCLIP · data curation · metadata balancing
Metadata checked 2026-07-28
2024CVPRPeer reviewed
Depth Anything: Unleashing the Power of Large-Scale Unlabeled Data
Lihe Yang et al.
L2–L3
3D / spatial vision Visual representations
3D / spatial vision · visual representation learning · monocular depth · unlabeled data · teacher-student learning
Official artifacts: Official code
Metadata checked 2026-08-08
2024NaturePeer reviewed
Detecting Hallucinations in Large Language Models Using Semantic Entropy
Sebastian Farquhar, Jannik Kossen, Lorenz Kuhn, and Yarin Gal
L2–L3
LLM evaluation / safety
LLM evaluation / safety · semantic entropy · hallucination detection · uncertainty
Official artifacts: Code
Metadata checked 2026-07-28
2024CVPRPeer reviewed
DETRs Beat YOLOs on Real-time Object Detection
Yian Zhao, Wenyu Lv, Shangliang Xu, Jinman Wei, Guanzhong Wang, Qingqing Dang, Yi Liu, and Jie Chen
L2–L3
Detection / segmentation
detection / segmentation · RT-DETR · real-time detection · set prediction · NMS-free detection
Official artifacts: Project website · Official implementation
Metadata checked 2026-08-10
2024TMLRPeer reviewed
DINOv2: Learning Robust Visual Features without Supervision
Maxime Oquab et al.
Research
Visual representations
self-supervised learning · DINOv2 · scaling · visual representation learning
Metadata checked 2026-07-27
2024NeurIPSPeer reviewed
Discrete Flow Matching
Itai Gat, Tal Remez, Neta Shaul, Felix Kreuk, Ricky T. Q. Chen, Gabriel Synnaeve, Yossi Adi, and Yaron Lipman
L2–L3
Generative / representation Generation / MT evaluation LLM pretraining
generative / representation learning · generation / translation evaluation · LLM pretraining / scaling · discrete flow matching · categorical probability paths · non-autoregressive generation
Metadata checked 2026-08-10
2024ICMLPeer reviewed
DistiLLM: Towards Streamlined Distillation for Large Language Models
Jongwoo Ko, Sungnyun Kim, Tianyi Chen, and Se-Young Yun
L2–L3
Peer-reviewed ICML 2024 proceedings paper.
Knowledge distillation Generation / MT evaluation LLM post-training LLM systems
knowledge distillation / model transfer · generation / translation evaluation · LLM post-training / alignment · LLM training / serving systems · DistiLLM · generative language distillation · skew KL · adaptive off-policy sampling · student trajectories
Official artifacts: Official code
Metadata checked 2026-08-23
2024OSDIPeer reviewed
DistServe: Disaggregating Prefill and Decoding for Goodput-Optimized Large Language Model Serving
Yinmin Zhong, Shengyu Liu, Junda Chen, Jianbo Hu, Yibo Zhu, Xuanzhe Liu, Xin Jin, and Hao Zhang
L2–L3
LLM systems
LLM training / serving systems · DistServe · disaggregated serving · prefill and decoding · goodput optimization
Official artifacts: Official implementation
Metadata checked 2026-08-10
2024ACLPeer reviewed
Dolma: An Open Corpus of Three Trillion Tokens for Language Model Pretraining Research
Luca Soldaini et al.
L2
LLM pretraining
LLM pretraining / scaling · Dolma · open corpus · data governance
Official artifacts: Dataset · Toolkit
Metadata checked 2026-07-28
2024CVPRPeer reviewed
DUSt3R: Geometric 3D Vision Made Easy
Shuzhe Wang et al.
L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · DUSt3R · point maps · 3D reconstruction · 3D / spatial vision
Metadata checked 2026-07-27
2024CVPRPeer reviewed
Florence-2: Advancing a Unified Representation for a Variety of Vision Tasks
Bin Xiao et al.
L2
Grounding / open vocabulary
grounding / open vocabulary · unified vision model · sequence interface · grounding
Metadata checked 2026-07-27
2024ICLRPeer reviewed
FROSTER: Frozen CLIP Is a Strong Teacher for Open-Vocabulary Action Recognition
Xiaohu Huang, Hao Zhou, Kun Yao, and Kai Han
L3
Published at ICLR 2024.
Knowledge distillation Image–text representation V&L evaluation / safety Video / 3D / embodied
knowledge distillation / model transfer · image–text representation · vision–language evaluation / safety · video / 3D / embodied vision · FROSTER · CLIP · open-vocabulary action recognition · residual feature distillation · video adaptation
Official artifacts: Official project · Official implementation
Metadata checked 2026-09-02
2024CVPRPeer reviewed
GLaMM: Pixel Grounding Large Multimodal Model
Hanoona Rasheed et al.
L2–L3
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · multimodal LLM · pixel grounding · region dialogue · grounded generation
Metadata checked 2026-07-27
2024ECCVPeer reviewed
Grounding DINO: Marrying DINO with Grounded Pre-Training for Open-Set Object Detection
Shilong Liu et al.
L3
Detection / segmentation Grounding / open vocabulary
detection / segmentation · Grounding DINO · grounded detection · open-set detection · grounding / open vocabulary
Metadata checked 2026-07-27
2024ECCVPeer reviewed
Grounding Image Matching in 3D with MASt3R
Vincent Leroy, Yohann Cabon, and Jérôme Revaud
L2–L3
3D / spatial vision
3D / spatial vision · MASt3R · 3D matching · pointmaps · image correspondence
Official artifacts: Official code · Official project article
Metadata checked 2026-08-10
2024CVPRPeer reviewed
HallusionBench: An Advanced Diagnostic Suite for Entangled Language Hallucination and Visual Illusion in Large Vision-Language Models
Tianrui Guan et al.
L2–L3
V&L evaluation / safety
vision–language evaluation / safety · visual illusion · language hallucination · diagnostic benchmark
Metadata checked 2026-07-27
2024ICMLPeer reviewed
HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
Mantas Mazeika et al.
L3
LLM evaluation / safety
LLM evaluation / safety · HarmBench · red teaming · robust refusal
Metadata checked 2026-07-27
2024CVPRPeer reviewed
Improved Baselines with Visual Instruction Tuning
Haotian Liu et al.
L2
Multimodal LLMs
multimodal LLM · LLaVA-1.5 · visual instruction tuning · connector
Metadata checked 2026-07-27
2024Transactions on Machine Learning ResearchPeer reviewed
Improving and Generalizing Flow-Based Generative Models with Minibatch Optimal Transport
Alexander Tong, Kilian Fatras, Nikolay Malkin, Guillaume Huguet, Yanlei Zhang, Jarrid Rector-Brooks, Guy Wolf, and Yoshua Bengio
L2–L3
Generative / representation Generative vision
generative / representation learning · generative vision · conditional flow matching · minibatch optimal transport · continuous normalizing flow · simulation-free training
Official artifacts: Official code
Metadata checked 2026-08-10
2024CVPRPeer reviewed
InternVL: Scaling up Vision Foundation Models and Aligning for Generic Visual-Linguistic Tasks
Zhe Chen et al.
L2–L3
Multimodal LLMs
multimodal LLM · large vision encoder · alignment
Metadata checked 2026-07-27
2024arXivPreprint
Jumping Ahead: Improving Reconstruction Fidelity with JumpReLU Sparse Autoencoders
Senthooran Rajamanoharan, Tom Lieberum, Nicolas Sonnerat, Arthur Conmy, Vikrant Varma, János Kramár, and Neel Nanda
L3
Interpretability
interpretability / attribution · SAE · JumpReLU · sparse autoencoders · learned thresholds · straight-through estimator · L0 sparsity
Official artifacts: Author PDF with implementation pseudocode
Metadata checked 2026-10-09
2024ICLRPeer reviewed
Kosmos-2: Grounding Multimodal Large Language Models to the World
Zhiliang Peng et al.
L2–L3
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · multimodal LLM · location tokens · grounded generation · referring
Metadata checked 2026-07-27
2024arXivPreprint
Learning biologically relevant features in a pathology foundation model using sparse autoencoders
Nhat Minh Le, Ciyue Shen, Neel Patel, Chintan Shah, Darpan Sanghavi, Blake Martin, Alfred Eng, Daniel Shenker, Harshith Padigela, Raymond Biju, Syed Ashar Javed, Jennifer Hipp, John Abel, Harsha Pokkalla, Sean Grullon, and Dinkar Juyal
L3
Review uses arXiv:2407.10785v3, including its supplement. A later NeurIPS 2025 Imageomics workshop paper has the title Evaluating the Utility of Sparse Autoencoders for Interpreting a Pathology Foundation Model; its inaccessible full PDF was not used to replace this version's claims or publication status.
Interpretability
evaluation · interpretability / attribution · SAE · sparse autoencoders · PLUTO · cell-count correlations · layer-wise representations · mechanistic interpretability
Official artifacts: Versioned full manuscript · Authors' later Imageomics workshop publication
Metadata checked 2026-10-09
2024ICLRPeer reviewed
Let’s Verify Step by Step
Hunter Lightman et al.
L3
LLM post-training Reasoning / verification
LLM post-training / alignment · process supervision · verifier · mathematical reasoning · LLM reasoning / verification
Official artifacts: PRM800K dataset
Metadata checked 2026-07-28
2024CVPRPeer reviewed
LISA: Reasoning Segmentation via Large Language Model
Xin Lai et al.
L2–L3
Grounding / open vocabulary Multimodal LLMs
grounding / open vocabulary · multimodal LLM · reasoning segmentation · segmentation token · mask decoder
Metadata checked 2026-07-27
2024ACLPeer reviewed
LongBench: A Bilingual, Multitask Benchmark for Long Context Understanding
Yushi Bai et al.
L2–L3
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · LongBench · long context · bilingual benchmark
Official artifacts: Benchmark
Metadata checked 2026-07-28
2024NaturePeer reviewed
Loss of Plasticity in Deep Continual Learning
Shibhansh Dohare et al.
L2–L3
Generative / representation Transfer / adaptation / continual
generative / representation learning · continual learning · plasticity · non-stationarity
Metadata checked 2026-07-27
2024TACLPeer reviewed
Lost in the Middle: How Language Models Use Long Contexts
Nelson F. Liu, Kevin Lin, John Hewitt, Ashwin Paranjape, Michele Bevilacqua, Fabio Petroni, and Percy Liang
L2
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · long context · position sensitivity · effective context
Official artifacts: Code and data
Metadata checked 2026-07-28
2024COLMPeer reviewed
Mamba: Linear-Time Sequence Modeling with Selective State Spaces
Albert Gu and Tri Dao
L3
Transformers / pretraining Neural networks
NLP transformers / pretraining · neural networks · Mamba · state space model · linear time
Metadata checked 2026-07-27
2024ICLRPeer reviewed
MathVista: Evaluating Mathematical Reasoning of Foundation Models in Visual Contexts
Pan Lu et al.
L3
V&L evaluation / safety
vision–language evaluation / safety · MathVista · visual mathematics · multimodal reasoning
Metadata checked 2026-07-27
2024ICLRPeer reviewed
MiniLLM: Knowledge Distillation of Large Language Models
Yuxian Gu, Li Dong, Furu Wei, and Minlie Huang
L2–L3
Knowledge distillation Generation / MT evaluation LLM post-training LLM systems
knowledge distillation / model transfer · generation / translation evaluation · LLM post-training / alignment · LLM training / serving systems · reverse KL · generative language models · policy gradient · teacher-mixed sampling
Official artifacts: Official code
Metadata checked 2026-08-10
2024arXivTechnical report
Mixtral of Experts
Albert Q. Jiang et al.
L2–L3
LLM pretraining
LLM pretraining / scaling · Mixtral · mixture of experts · open weights
Metadata checked 2026-07-27
2024ECCVPeer reviewed
MM-SafetyBench: A Benchmark for Safety Evaluation of Multimodal Large Language Models
Xin Liu et al.
L2–L3
V&L evaluation / safety
vision–language evaluation / safety · multimodal safety · image-based attack · jailbreak
Metadata checked 2026-07-27
2024ICMLPeer reviewed
MM-Vet: Evaluating Large Multimodal Models for Integrated Capabilities
Weihao Yu et al.
L2–L3
V&L evaluation / safety
vision–language evaluation / safety · integrated capabilities · open-ended evaluation · LLM judge
Metadata checked 2026-07-27
2024ECCVPeer reviewed
MM1: Methods, Analysis & Insights from Multimodal LLM Pre-training
Brandon McKinzie et al.
L2–L3
Multimodal LLMs
multimodal LLM · multimodal pretraining · data mixture · architecture ablation
Metadata checked 2026-07-28
2024ECCVPeer reviewed
MMBench: Is Your Multi-modal Model an All-around Player?
Yuan Liu et al.
L2
V&L evaluation / safety
vision–language evaluation / safety · MLLM benchmark · circular evaluation · capability taxonomy
Metadata checked 2026-07-28
2024CVPRPeer reviewed
MMMU: A Massive Multi-discipline Multimodal Understanding and Reasoning Benchmark for Expert AGI
Xiang Yue et al.
L2–L3
V&L evaluation / safety
vision–language evaluation / safety · multidiscipline benchmark · expert reasoning · MLLM
Metadata checked 2026-07-27
2024CVPRPeer reviewed
MobileCLIP: Fast Image-Text Models through Multi-Modal Reinforced Training
Pavan Kumar Anasosalu Vasu et al.
L2
Image–text representation Knowledge distillation
image–text representation · efficient encoder · distillation · reinforced dataset · knowledge distillation / model transfer
Metadata checked 2026-07-27
2024ICMLPeer reviewed
Model Alignment as Prospect Theoretic Optimization
Kawin Ethayarajh, Winnie Xu, Niklas Muennighoff, Dan Jurafsky, and Douwe Kiela
L2–L3
Loss / objectives LLM post-training
loss / objective design · LLM post-training / alignment · KTO · prospect theory · human-aware loss · binary feedback · language-model alignment
Official artifacts: Official HALOs code
Metadata checked 2026-08-10
2024TMLRPeer reviewed
Multimodal Chain-of-Thought Reasoning in Language Models
Zhuosheng Zhang et al.
L2
Multimodal LLMs
multimodal LLM · multimodal reasoning · chain of thought · rationales
Metadata checked 2026-07-27
2024ACLPeer reviewed
OLMo: Accelerating the Science of Language Models
Dirk Groeneveld et al.
L2–L3
LLM pretraining
LLM pretraining / scaling · OLMo · open data · open training
Metadata checked 2026-07-27
2024ICLRPeer reviewed
On-Policy Distillation of Language Models: Learning from Self-Generated Mistakes
Rishabh Agarwal, Nino Vieillard, Yongchao Zhou, Piotr Stanczyk, Sabela Ramos, Matthieu Geist, and Olivier Bachem
L2–L3
Knowledge distillation Generation / MT evaluation LLM post-training
knowledge distillation / model transfer · generation / translation evaluation · LLM post-training / alignment · on-policy distillation · generalized knowledge distillation · autoregressive language models · student rollouts · Jensen-Shannon divergence
Metadata checked 2026-08-10
2024NeurIPS Datasets and BenchmarksPeer reviewed
OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments
Tianbao Xie et al.
L2–L3
The peer-reviewed version of record was published in the NeurIPS 2024 Datasets and Benchmarks Track.
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · computer-use agent · environment benchmark · multimodal agent
Metadata checked 2026-07-28
2024arXivPreprint
PaliGemma: A versatile 3B VLM for transfer
Lucas Beyer et al.
L2
Multimodal LLMs
multimodal LLM · compact VLM · transfer · open weights
Metadata checked 2026-07-27
2024arXivTechnical report
Pixtral 12B
Pravesh Agrawal et al.
L2
Multimodal LLMs
multimodal LLM · variable resolution · multimodal decoder · open weights
Metadata checked 2026-07-27
2024arXivPreprint
Qwen2-VL: Enhancing Vision-Language Model's Perception of the World at Any Resolution
Peng Wang et al.
L2–L3
Multimodal LLMs Captioning / VQA Grounding / open vocabulary Generation / video / agents
multimodal LLM · captioning / VQA · dynamic resolution · multimodal RoPE · image and video · grounding / open vocabulary · multimodal generation / agents
Metadata checked 2026-07-27
2024ICLRPeer reviewed
RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval
Parth Sarthi et al.
L3
RAG / tools / agents
RAG / tools / agents · RAPTOR · hierarchical retrieval · summarization
Metadata checked 2026-07-27
2024TMLRPeer reviewed
Revisiting Feature Prediction for Learning Visual Representations from Video
Adrien Bardes et al.
L3
Video / 3D / embodied
video / 3D / embodied vision · V-JEPA · feature prediction · video SSL
Metadata checked 2026-07-27
2024NeurocomputingPeer reviewed
RoFormer: Enhanced Transformer with Rotary Position Embedding
Jianlin Su et al.
L2
First released as an arXiv preprint in 2021; the catalog records the 2024 Neurocomputing version of record.
Transformers / pretraining LLM pretraining
NLP transformers / pretraining · LLM pretraining / scaling · RoPE · position encoding · Transformer
Metadata checked 2026-07-28
2024NaturePeer reviewed
Scaling neural machine translation to 200 languages
Marta R. Costa-jussà et al.
L2–L3
Multilingual / low resource Generation / MT evaluation
multilingual / low-resource NLP · generation / translation evaluation · NLLB · multilingual translation · low-resource languages
Official artifacts: Official NLLB artifacts
Metadata checked 2026-08-08
2024ICMLPeer reviewed
Scaling Rectified Flow Transformers for High-Resolution Image Synthesis
Patrick Esser et al.
L2–L3
Generative / representation Generative vision Generation / video / agents
generative / representation learning · generative vision · multimodal generation / agents · rectified flow · multimodal diffusion transformer · text-to-image synthesis
Metadata checked 2026-08-10
2024CVPRPeer reviewed
SEED-Bench: Benchmarking Multimodal Large Language Models
Bohao Li et al.
L2–L3
V&L evaluation / safety
vision–language evaluation / safety · SEED-Bench · multimodal evaluation · generative comprehension
Metadata checked 2026-07-27
2024ICLRPeer reviewed
Self-RAG: Learning to Retrieve, Generate, and Critique through Self-Reflection
Akari Asai et al.
L3
RAG / tools / agents
RAG / tools / agents · Self-RAG · retrieval · self-critique
Metadata checked 2026-07-27
2024NeurIPSPeer reviewed
SGLang: Efficient Execution of Structured Language Model Programs
Lianmin Zheng, Liangsheng Yin, Zhiqiang Xie, Chuyue Sun, Jeff Huang, Cody Hao Yu, Shiyi Cao, Christos Kozyrakis, Ion Stoica, Joseph E. Gonzalez, Clark Barrett, and Ying Sheng
L2–L3
LLM systems
LLM training / serving systems · SGLang · structured language model programs · RadixAttention · KV cache reuse
Official artifacts: Official implementation
Metadata checked 2026-08-10
2024NeurIPSPeer reviewed
SimPO: Simple Preference Optimization with a Reference-Free Reward
Yu Meng, Mengzhou Xia, and Danqi Chen
L2–L3
Loss / objectives LLM post-training
loss / objective design · LLM post-training / alignment · SimPO · preference optimization · reference-free alignment · length-normalized reward · target reward margin
Official artifacts: Official code and models
Metadata checked 2026-08-10
2024MICCAIPeer reviewed
SlideGCD: Slide-based Graph Collaborative Training with Knowledge Distillation for Whole Slide Image Classification
Tong Shu, Jun Shi, Dongdong Sun, Zhiguo Jiang, and Yushan Zheng
L3
Knowledge distillation
knowledge distillation / model transfer · MIL / WSI · whole-slide image · multiple-instance learning · slide graph · online distillation
Official artifacts: Official code
Metadata checked 2026-08-10
2024ICLRPeer reviewed
Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training
Hong Liu et al.
L3
Neural networks
neural networks · Sophia · second-order optimization · language-model training
Metadata checked 2026-07-27
2024ICLRPeer reviewed
Sparse Autoencoders Find Highly Interpretable Features in Language Models
Hoagy Cunningham, Aidan Ewart, Logan Riggs Smith, Robert Huben, and Lee Sharkey
L2–L3
LLM interpretability Interpretability
LLM interpretability / editing / unlearning · sparse autoencoders · feature dictionaries · monosemanticity · interpretability / attribution
Metadata checked 2026-08-08
2024ICLRPeer reviewed
SWE-bench: Can Language Models Resolve Real-World GitHub Issues?
Carlos E. Jimenez et al.
L3
RAG / tools / agents LLM evaluation / safety Code / software engineering
RAG / tools / agents · LLM evaluation / safety · SWE-bench · software agent · repository-level evaluation · code generation / software engineering
Official artifacts: Benchmark
Metadata checked 2026-07-28
2024NeurIPSPeer reviewed
The Road Less Scheduled
Aaron Defazio et al.
L3
Neural networks
neural networks · schedule-free optimization · learning-rate schedules
Metadata checked 2026-07-27
2024ICMLPeer reviewed
The WMDP Benchmark: Measuring and Reducing Malicious Use with Unlearning
Nathaniel Li, Alexander Pan, Anjali Gopal, Summer Yue, Daniel Berrios, Alice Gatti, Justin D. Li, Ann-Kathrin Dombrowski, Shashwat Goel, Gabriel Mukobi, Nathan Helm-Burger, Rassin Lababidi, Lennart Justen, Andrew B. Liu, Michael Chen, Isabelle Barrass, Oliver Zhang, Xiaoyuan Zhu, Rishub Tamirisa, Bhrugu Bharathi, Ariel Herbert-Voss, Cort B. Breuer, Andy Zou, Mantas Mazeika, Zifan Wang, Palash Oswal, Weiran Lin, Adam A. Hunt, Justin Tienken-Harder, Kevin Y. Shih, Kemper Talley, John Guan, Ian Steneker, David Campbell, Brad Jokubaitis, Steven Basart, Stephen Fitz, Ponnurangam Kumaraguru, Kallol Krishna Karmakar, Uday Tupakula, Vijay Varadharajan, Yan Shoshitaishvili, Jimmy Ba, Kevin M. Esvelt, Alexandr Wang, and Dan Hendrycks
L2–L3
LLM interpretability LLM evaluation / safety
LLM interpretability / editing / unlearning · LLM evaluation / safety · WMDP · machine unlearning · representation misdirection · hazardous knowledge evaluation
Official artifacts: Official benchmark and RMU code · Project website
Metadata checked 2026-08-10
2024ICLRPeer reviewed
ToolLLM: Facilitating Large Language Models to Master 16000+ Real-world APIs
Yujia Qin et al.
L2–L3
RAG / tools / agents
RAG / tools / agents · ToolLLM · tool learning · API use
Official artifacts: ToolBench
Metadata checked 2026-07-28
2024ICMLPeer reviewed
Transformers are SSMs: Generalized Models and Efficient Algorithms Through Structured State Space Duality
Tri Dao and Albert Gu
L3
Neural networks Transformers / pretraining LLM pretraining
neural networks · Mamba-2 · state space model · structured state space duality · NLP transformers / pretraining · LLM pretraining / scaling
Metadata checked 2026-07-27
2024ACLPeer reviewed
Video-ChatGPT: Towards Detailed Video Understanding via Large Vision and Language Models
Muhammad Maaz et al.
L2
Generation / video / agents
multimodal generation / agents · video conversation · instruction tuning · evaluation
Metadata checked 2026-07-27
2024ICMLPeer reviewed
VideoPoet: A Large Language Model for Zero-Shot Video Generation
Dan Kondratyuk et al.
L2–L3
Generation / video / agents
multimodal generation / agents · video generation · multimodal tokens · zero-shot composition
Metadata checked 2026-07-28
2024CVPRPeer reviewed
VILA: On Pre-training for Visual Language Models
Ji Lin et al.
L2
Multimodal LLMs
multimodal LLM · visual-language pretraining · data mixture · interleaved data
Metadata checked 2026-07-28
2024ICLRPeer reviewed
Vision Transformers Need Registers
Timothée Darcet, Maxime Oquab, Julien Mairal, and Piotr Bojanowski
L2–L3
Published as an ICLR 2024 oral paper.
Visual representations
visual representation learning · register tokens · feature artifacts · dense representations
Official artifacts: DINOv2 implementation
Metadata checked 2026-08-08
2024NeurIPSPeer reviewed
Visual Autoregressive Modeling: Scalable Image Generation via Next-Scale Prediction
Keyu Tian et al.
L3
Generative vision
generative vision · VAR · autoregressive generation · next-scale prediction
Metadata checked 2026-07-27
2024ACLPeer reviewed
VisualWebArena: Evaluating Multimodal Agents on Realistic Visual Web Tasks
Jing Yu Koh et al.
L2
The peer-reviewed version of record was published at ACL 2024.
Generation / video / agents V&L evaluation / safety
multimodal generation / agents · vision–language evaluation / safety · web agent · visual grounding · interactive benchmark
Metadata checked 2026-07-28
2024NeurIPSPeer reviewed
VMamba: Visual State Space Model
Yue Liu et al.
L3
Visual representations
visual representation learning · VMamba · state-space model · visual backbone
Metadata checked 2026-07-27
2024NeurIPSPeer reviewed
Wasserstein Distance Rivals Kullback-Leibler Divergence for Knowledge Distillation
Jiaming Lv, Haoyuan Yang, and Peihua Li
L2–L3
Knowledge distillation Loss / objectives
knowledge distillation / model transfer · loss / objective design · optimal transport · Wasserstein distance · logit distillation · feature distillation · Bures metric
Official artifacts: Official project page · Official code
Metadata checked 2026-08-10
2024ICLRPeer reviewed
WebArena: A Realistic Web Environment for Building Autonomous Agents
Shuyan Zhou et al.
L3
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · WebArena · web agent · interactive evaluation
Official artifacts: Environment and benchmark
Metadata checked 2026-07-28
2024NeurIPSPeer reviewed
xLSTM: Extended Long Short-Term Memory
Maximilian Beck et al.
L3
Neural networks
neural networks · xLSTM · recurrent neural network · matrix memory
Metadata checked 2026-07-27
2023ACM Transactions on Graphics / SIGGRAPHPeer reviewed
3D Gaussian Splatting for Real-Time Radiance Field Rendering
Bernhard Kerbl et al.
L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · 3D Gaussian splatting · neural rendering · radiance field · 3D / spatial vision
Metadata checked 2026-07-27
2023Foundations and Trends in Machine LearningPeer reviewed
A Gentle Introduction to Conformal Prediction and Distribution-Free Uncertainty Quantification
Anastasios N. Angelopoulos and Stephen Bates
L2–L3
Evaluation / reproducibility
evaluation / reproducibility · conformal prediction · coverage · distribution-free uncertainty
Metadata checked 2026-07-27
2023ICCVPeer reviewed
Adding Conditional Control to Text-to-Image Diffusion Models
Lvmin Zhang, Anyi Rao, and Maneesh Agrawala
L2
Generative vision Generation / video / agents
generative vision · ControlNet · spatial conditioning · diffusion models · multimodal generation / agents
Official artifacts: Code
Metadata checked 2026-07-28
2023NIST AI 100-1Guideline
Artificial Intelligence Risk Management Framework (AI RMF 1.0)
Elham Tabassi
L1–L2
NIST states that AI RMF 1.0 is being revised; this record points to the stable framework landing page.
Evaluation / reproducibility LLM evaluation / safety
evaluation / reproducibility · AI risk management · governance · trustworthiness · LLM evaluation / safety
Metadata checked 2026-07-27
2023JMLRPeer reviewed
Atlas: Few-shot Learning with Retrieval Augmented Language Models
Gautier Izacard et al.
L2–L3
RAG / tools / agents
RAG / tools / agents · Atlas · few-shot learning · retrieval augmentation
Metadata checked 2026-07-27
2023EUSIPCOPeer reviewed
Attention to Detail: Inter-Resolution Knowledge Distillation
Rocío del Amor, Julio Silva-Rodríguez, Adrián Colomer, and Valery Naranjo
L3
Peer-reviewed EUSIPCO 2023 proceedings paper, pages 985–989.
Knowledge distillation
knowledge distillation / model transfer · WSI engineering · cell / tissue modeling · inter-resolution distillation · Grad-CAM · attention transfer · prostate grading · low-resolution deployment
Official artifacts: Open proceedings paper · Official code
Metadata checked 2026-08-23
2023TMLRPeer reviewed
Beyond the Imitation Game: Quantifying and Extrapolating the Capabilities of Language Models
BIG-bench collaboration
L2
LLM evaluation / safety
LLM evaluation / safety · BIG-bench · capability evaluation · scaling
Metadata checked 2026-07-27
2023ICMLPeer reviewed
BLIP-2: Bootstrapping Language-Image Pre-training with Frozen Image Encoders and Large Language Models
Junnan Li et al.
L2–L3
Multimodal LLMs Captioning / VQA
multimodal LLM · multimodal bridge · frozen models · captioning / VQA
Metadata checked 2026-07-27
2023arXivPreprint
Code Llama: Open Foundation Models for Code
Baptiste Rozière, Jonas Gehring, Fabian Gloeckle, Sten Sootla, Itai Gat, Xiaoqing Ellen Tan, Yossi Adi, Jingyu Liu, Romain Sauvestre, Tal Remez, Jérémy Rapin, Artyom Kozhevnikov, Ivan Evtimov, Joanna Bitton, Manish Bhatt, Cristian Canton Ferrer, Aaron Grattafiori, Wenhan Xiong, Alexandre Défossez, Jade Copet, Faisal Azhar, Hugo Touvron, Louis Martin, Nicolas Usunier, Thomas Scialom, and Gabriel Synnaeve
L2–L3
Code / software engineering
code generation / software engineering · Code Llama · code foundation model · infilling · long context · instruction tuning
Official artifacts: Official implementation and model documentation
Metadata checked 2026-08-20
2023ICMLPeer reviewed
Consistency Models
Yang Song et al.
L3
Generative / representation Generative vision
generative / representation learning · consistency model · one-step generation · distillation · generative vision
Metadata checked 2026-07-27
2023NeurIPS Datasets and BenchmarksPeer reviewed
DataComp: In Search of the Next Generation of Multimodal Datasets
Samir Yitzhak Gadre et al.
L2
Image–text representation V&L evaluation / safety Multimodal retrieval
image–text representation · vision–language evaluation / safety · data curation · benchmark · fixed compute · multimodal embedding / retrieval
Metadata checked 2026-07-27
2023ICLRPeer reviewed
Diffusion Posterior Sampling for General Noisy Inverse Problems
Hyungjin Chung, Jeongsol Kim, Michael T. McCann, Marc L. Klasky, and Jong Chul Ye
L2–L3
Restoration / inverse problems
DPS · diffusion posterior sampling · noisy nonlinear inverse problems · likelihood guidance
Official artifacts: Official implementation · arXiv record
Metadata checked 2026-08-20
2023ICLRPeer reviewed
DINO: DETR with Improved DeNoising Anchor Boxes for End-to-End Object Detection
Hao Zhang et al.
L3
Detection / segmentation
detection / segmentation · DINO detector · DETR · denoising training
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
Direct Preference Optimization: Your Language Model is Secretly a Reward Model
Rafael Rafailov et al.
L2–L3
LLM post-training Loss / objectives
LLM post-training / alignment · DPO · preference optimization · alignment · loss / objective design
Metadata checked 2026-07-27
2023Findings of ACLPeer reviewed
Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizes
Cheng-Yu Hsieh, Chun-Liang Li, Chih-Kuan Yeh, Hootan Nakhost, Yasuhisa Fujii, Alexander Ratner, Ranjay Krishna, Chen-Yu Lee, and Tomas Pfister
L2–L3
Knowledge distillation Reasoning / verification
knowledge distillation / model transfer · LLM reasoning / verification · rationale distillation · chain of thought · multitask learning
Official artifacts: Official code
Metadata checked 2026-08-10
2023CoRL 2022 / PMLRPeer reviewed
Do As I Can, Not As I Say: Grounding Language in Robotic Affordances
Brian Ichter et al.
L2–L3
The conference was held in 2022; the PMLR version of record was published in 2023.
Generation / video / agents
multimodal generation / agents · SayCan · robotics · affordance grounding · language planning
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
DoReMi: Optimizing Data Mixtures Speeds Up Language Model Pretraining
Sang Michael Xie, Shibani Santurkar, Tengyu Ma, and Percy Liang
L2–L3
LLM pretraining
LLM pretraining / scaling · DoReMi · data mixture · domain reweighting
Official artifacts: Code
Metadata checked 2026-07-28
2023CVPRPeer reviewed
DreamBooth: Fine Tuning Text-to-Image Diffusion Models for Subject-Driven Generation
Nataniel Ruiz et al.
L2
Generation / video / agents
multimodal generation / agents · subject-driven generation · personalization · prior preservation
Metadata checked 2026-07-28
2023SOSPPeer reviewed
Efficient Memory Management for Large Language Model Serving with PagedAttention
Woosuk Kwon et al.
L2–L3
LLM evaluation / safety LLM systems
LLM evaluation / safety · vLLM · PagedAttention · LLM serving · LLM training / serving systems
Official artifacts: Serving system
Metadata checked 2026-07-28
2023EMNLPPeer reviewed
Enabling Large Language Models to Generate Text with Citations
Tianyu Gao et al.
L2–L3
RAG / tools / agents LLM evaluation / safety
RAG / tools / agents · LLM evaluation / safety · ALCE · citation generation · attribution evaluation
Official artifacts: Code and data
Metadata checked 2026-07-28
2023EMNLPPeer reviewed
Evaluating Object Hallucination in Large Vision-Language Models
Yifan Li et al.
L2
V&L evaluation / safety
vision–language evaluation / safety · POPE · object hallucination · polling evaluation
Metadata checked 2026-07-27
2023ICMLPeer reviewed
Fast Inference from Transformers via Speculative Decoding
Yaniv Leviathan, Matan Kalman, and Yossi Matias
L2–L3
LLM evaluation / safety LLM systems
LLM evaluation / safety · speculative decoding · inference latency · exact sampling · LLM training / serving systems
Metadata checked 2026-07-28
2023ICLRPeer reviewed
Flow Matching for Generative Modeling
Yaron Lipman et al.
L3
Generative / representation Generative vision
generative / representation learning · flow matching · continuous normalizing flow · optimal transport · generative vision
Metadata checked 2026-07-27
2023ICLRPeer reviewed
Flow Straight and Fast: Learning to Generate and Transfer Data with Rectified Flow
Xingchao Liu, Chengyue Gong, and Qiang Liu
L3
Generative / representation
generative / representation learning · rectified flow · transport · generative modeling
Metadata checked 2026-07-27
2023EMNLPPeer reviewed
G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment
Yang Liu, Dan Iter, Yichong Xu, Shuohang Wang, Ruochen Xu, and Chenguang Zhu
L2–L3
Generation / MT evaluation
generation / translation evaluation · G-Eval · LLM-as-a-judge · NLG evaluation · human alignment
Official artifacts: Official implementation
Metadata checked 2026-08-10
2023NeurIPS Datasets and BenchmarksPeer reviewed
GenEval: An Object-Focused Framework for Evaluating Text-to-Image Alignment
Dhruba Ghosh, Hannaneh Hajishirzi, and Ludwig Schmidt
L2
V&L evaluation / safety Generation / video / agents
vision–language evaluation / safety · multimodal generation / agents · text-to-image evaluation · object composition · prompt alignment
Metadata checked 2026-07-28
2023EMNLPPeer reviewed
GQA: Training Generalized Multi-Query Transformer Models from Multi-Head Checkpoints
Joshua Ainslie et al.
L2–L3
Transformers / pretraining LLM pretraining
NLP transformers / pretraining · LLM pretraining / scaling · grouped-query attention · KV cache · inference
Metadata checked 2026-07-27
2023TMLRPeer reviewed
Holistic Evaluation of Language Models
Percy Liang et al.
L2
LLM evaluation / safety
LLM evaluation / safety · HELM · holistic evaluation · transparency
Metadata checked 2026-07-27
2023CVPRPeer reviewed
Image as a Foreign Language: BEiT Pretraining for Vision and Vision-Language Tasks
Wenhui Wang et al.
L2
Image–text representation Captioning / VQA
image–text representation · captioning / VQA · multiway transformer · masked data modelling · shared backbone
Metadata checked 2026-07-28
2023CVPRPeer reviewed
ImageBind: One Embedding Space To Bind Them All
Rohit Girdhar et al.
L2
Image–text representation Multimodal retrieval
image–text representation · multimodal embedding · six modalities · alignment · multimodal embedding / retrieval
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
InstructBLIP: Towards General-purpose Vision-Language Models with Instruction Tuning
Wenliang Dai et al.
L2
Multimodal LLMs
multimodal LLM · instruction tuning · Q-Former · frozen models
Metadata checked 2026-07-27
2023CVPRPeer reviewed
InstructPix2Pix: Learning to Follow Image Editing Instructions
Tim Brooks, Aleksander Holynski, and Alexei A. Efros
L2
Generative vision Generation / video / agents
generative vision · instruction-based editing · image editing · diffusion models · multimodal generation / agents
Official artifacts: Code
Metadata checked 2026-07-28
2023NeurIPS Datasets and BenchmarksPeer reviewed
Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena
Lianmin Zheng et al.
L2–L3
LLM evaluation / safety
LLM evaluation / safety · MT-Bench · Chatbot Arena · model-as-judge
Metadata checked 2026-07-27
2023IEEE Journal of Biomedical and Health InformaticsPeer reviewed
Knowledge Distillation in Histology Landscape by Multi-Layer Features Supervision
Sajid Javed, Arif Mahmood, Talha Qaiser, and Naoufel Werghi
L3
Peer-reviewed journal article 27(4):2037–2046.
Knowledge distillation
knowledge distillation / model transfer · cell / tissue modeling · KDTP · multi-layer feature supervision · attention weighting · histology phenotype classification · student compression
Official artifacts: Open author manuscript
Metadata checked 2026-08-23
2023NeurIPSPeer reviewed
Language Is Not All You Need: Aligning Perception with Language Models
Shaohan Huang et al.
L2
The peer-reviewed version of record was published in the NeurIPS 2023 Main Conference Track.
Multimodal LLMs
multimodal LLM · Kosmos-1 · multimodal language model · in-context learning
Metadata checked 2026-07-28
2023WACVPeer reviewed
Large-to-Small Image Resolution Asymmetry in Deep Metric Learning
Pavel Suma and Giorgos Tolias
L2–L3
Peer-reviewed WACV 2023 proceedings paper.
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · resolution asymmetry · deep metric learning · cross-resolution retrieval · teacher–student training · low-resolution query
Official artifacts: Official code
Metadata checked 2026-08-23
2023arXivTechnical report
LLaMA: Open and Efficient Foundation Language Models
Hugo Touvron et al.
L2
LLM pretraining
LLM pretraining / scaling · LLaMA · foundation language model · data scaling
Metadata checked 2026-07-27
2023arXivPreprint
LongNet: Scaling Transformers to 1,000,000,000 Tokens
Jiayu Ding, Shuming Ma, Li Dong, Xingxing Zhang, Shaohan Huang, Wenhui Wang, Nanning Zheng, and Furu Wei
L2–L3
RAG / tools / agents
RAG / tools / agents · LongNet · dilated attention · sparse attention · long context · distributed training
Official artifacts: Microsoft Research publication page · Official TorchScale implementation
Metadata checked 2026-08-20
2023ICLRPeer reviewed
Mass-Editing Memory in a Transformer
Kevin Meng, Arnab Sen Sharma, Alex Andonian, Yonatan Belinkov, and David Bau
L2–L3
LLM interpretability
LLM interpretability / editing / unlearning · MEMIT · mass model editing · transformer memory · factual associations
Official artifacts: Official code
Metadata checked 2026-08-08
2023NeurIPSPeer reviewed
MEGABYTE: Predicting Million-byte Sequences with Multiscale Transformers
Lili Yu, Daniel Simig, Colin Flaherty, Armen Aghajanyan, Luke Zettlemoyer, and Mike Lewis
L2–L3
Text / sequence models Transformers / pretraining LLM pretraining
text / sequence modeling · NLP transformers / pretraining · LLM pretraining / scaling · MEGABYTE · byte-level modeling · multiscale Transformer · hierarchical patches
Metadata checked 2026-08-10
2023TACLPeer reviewed
MIRACL: A Multilingual Retrieval Dataset Covering 18 Diverse Languages
Xinyu Zhang et al.
L2
NLP tasks / evaluation Multilingual / low resource
NLP tasks / evaluation · MIRACL · multilingual retrieval · benchmark · multilingual / low-resource NLP
Official artifacts: Dataset and evaluation
Metadata checked 2026-07-28
2023ICCVPeer reviewed
Multiple Instance Learning Framework with Masked Hard Instance Mining for Whole Slide Image Classification
Wenhao Tang et al.
L3
Knowledge distillation
MIL / WSI · MHIM-MIL · hard-instance mining · teacher–student learning · knowledge distillation / model transfer
Metadata checked 2026-07-27
2023CVPRPeer reviewed
Open-Vocabulary Panoptic Segmentation with Text-to-Image Diffusion Models
Jiarui Xu et al.
L2–L3
Grounding / open vocabulary Detection / segmentation
grounding / open vocabulary · open-vocabulary segmentation · diffusion features · panoptic · detection / segmentation
Metadata checked 2026-07-27
2023ICLRPeer reviewed
PaLI: A Jointly-Scaled Multilingual Language-Image Model
Xi Chen et al.
L2
Multimodal LLMs Captioning / VQA
multimodal LLM · captioning / VQA · multilingual MLLM · joint scaling · multi-task transfer
Metadata checked 2026-07-28
2023ICMLPeer reviewed
PaLM-E: An Embodied Multimodal Language Model
Danny Driess et al.
L2–L3
Generation / video / agents Video / 3D / embodied Multimodal LLMs
multimodal generation / agents · embodied language model · sensor tokens · robotics · video / 3D / embodied vision · multimodal LLM
Metadata checked 2026-07-27
2023JMLRPeer reviewed
PaLM: Scaling Language Modeling with Pathways
Aakanksha Chowdhery et al.
L2–L3
LLM pretraining
LLM pretraining / scaling · PaLM · scaling · Pathways
Metadata checked 2026-07-27
2023ICLRPeer reviewed
Phenaki: Variable Length Video Generation from Open Domain Textual Descriptions
Ruben Villegas et al.
L2–L3
Generation / video / agents
multimodal generation / agents · text-to-video · causal video tokens · variable-length generation
Metadata checked 2026-07-28
2023ICMLPeer reviewed
Pix2Struct: Screenshot Parsing as Pretraining for Visual Language Understanding
Kenton Lee et al.
L2–L3
Document intelligence Captioning / VQA
document intelligence · captioning / VQA · Pix2Struct · screenshot parsing · variable-resolution input
Metadata checked 2026-08-08
2023ACLPeer reviewed
Precise Zero-Shot Dense Retrieval without Relevance Labels
Luyu Gao, Xueguang Ma, Jimmy Lin, and Jamie Callan
L2–L3
NLP tasks / evaluation RAG / tools / agents
NLP tasks / evaluation · RAG / tools / agents · HyDE · zero-shot dense retrieval · hypothetical documents · retrieval augmentation
Official artifacts: Official implementation
Metadata checked 2026-08-10
2023NeurIPSPeer reviewed
QLoRA: Efficient Finetuning of Quantized LLMs
Tim Dettmers et al.
L2–L3
Transformers / pretraining LLM post-training Transfer / adaptation / continual
NLP transformers / pretraining · LLM post-training / alignment · QLoRA · quantization · parameter-efficient tuning
Metadata checked 2026-07-27
2023arXivPreprint
Qwen-VL: A Versatile Vision-Language Model for Understanding, Localization, Text Reading, and Beyond
Jinze Bai et al.
L2–L3
Multimodal LLMs
multimodal LLM · multilingual · OCR · grounding · visual dialogue
Metadata checked 2026-07-27
2023ICLRPeer reviewed
ReAct: Synergizing Reasoning and Acting in Language Models
Shunyu Yao et al.
L2
LLM post-training RAG / tools / agents
LLM post-training / alignment · RAG / tools / agents · ReAct · reasoning · tool use
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
Reflexion: Language Agents with Verbal Reinforcement Learning
Noah Shinn, Federico Cassano, Ashwin Gopinath, Karthik Narasimhan, and Shunyu Yao
L2–L3
RAG / tools / agents
RAG / tools / agents · Reflexion · language agents · verbal reinforcement · episodic memory
Official artifacts: Official implementation and data
Metadata checked 2026-08-10
2023Robotics: Science and SystemsPeer reviewed
RT-1: Robotics Transformer for Real-World Control at Scale
Anthony Brohan et al.
L2
Generation / video / agents Video / 3D / embodied
multimodal generation / agents · robotics transformer · action tokenization · real-world control · video / 3D / embodied vision
Metadata checked 2026-07-28
2023CoRLPeer reviewed
RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control
Brianna Zitkovich et al.
L2–L3
Generation / video / agents Video / 3D / embodied
multimodal generation / agents · vision-language-action · robot control · action tokens · video / 3D / embodied vision
Metadata checked 2026-07-27
2023ICCVPeer reviewed
Scalable Diffusion Models with Transformers
William Peebles and Saining Xie
L3
Generative vision Generation / video / agents
generative vision · DiT · diffusion transformer · scaling · multimodal generation / agents
Metadata checked 2026-07-27
2023ICMLPeer reviewed
Scaling Laws for Reward Model Overoptimization
Leo Gao, John Schulman, and Jacob Hilton
L2–L3
LLM post-training
LLM post-training / alignment · reward model · overoptimization · distribution shift
Metadata checked 2026-07-28
2023ICCVPeer reviewed
Segment Anything
Alexander Kirillov et al.
L2–L3
Detection / segmentation Grounding / open vocabulary
detection / segmentation · SAM · promptable segmentation · foundation model · grounding / open vocabulary
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
Segment Everything Everywhere All at Once
Xueyan Zou et al.
L2
Grounding / open vocabulary
grounding / open vocabulary · SEEM · promptable segmentation · open vocabulary
Metadata checked 2026-07-27
2023ICLRPeer reviewed
Self-Consistency Improves Chain of Thought Reasoning in Language Models
Xuezhi Wang et al.
L2
LLM post-training Reasoning / verification
LLM post-training / alignment · self-consistency · chain of thought · test-time sampling · LLM reasoning / verification
Metadata checked 2026-07-28
2023ACLPeer reviewed
Self-Instruct: Aligning Language Models with Self-Generated Instructions
Yizhong Wang et al.
L2
LLM post-training
LLM post-training / alignment · Self-Instruct · synthetic instructions · instruction tuning
Metadata checked 2026-07-27
2023CVPRPeer reviewed
Self-Supervised Learning from Images with a Joint-Embedding Predictive Architecture
Mahmoud Assran et al.
Research
Generative / representation Visual representations
self-supervised learning · I-JEPA · joint-embedding prediction · generative / representation learning · visual representation learning
Metadata checked 2026-07-27
2023ICCVPeer reviewed
Sigmoid Loss for Language Image Pre-Training
Xiaohua Zhai et al.
L2
Image–text representation Detection / segmentation Visual representations Multimodal retrieval Loss / objectives
image–text representation · pathology VLM · SigLIP · image–text alignment · detection / segmentation · visual representation learning · multimodal embedding / retrieval · loss / objective design
Metadata checked 2026-07-27
2023ICMLPeer reviewed
SmoothQuant: Accurate and Efficient Post-Training Quantization for Large Language Models
Guangxuan Xiao et al.
L2–L3
LLM evaluation / safety LLM systems
LLM evaluation / safety · SmoothQuant · post-training quantization · inference efficiency · LLM training / serving systems
Official artifacts: Code
Metadata checked 2026-07-28
2023NeurIPS Datasets and BenchmarksPeer reviewed
SugarCrepe: Fixing Hackable Benchmarks for Vision-Language Compositionality
Cheng-Yu Hsieh et al.
L2–L3
V&L evaluation / safety
vision–language evaluation / safety · compositionality · hard negatives · benchmark bias
Metadata checked 2026-07-28
2023NeurIPS Datasets and BenchmarksPeer reviewed
The RefinedWeb Dataset for Falcon LLM: Outperforming Curated Corpora with Web Data, and Web Data Only
Guilherme Penedo et al.
L2
LLM pretraining
LLM pretraining / scaling · RefinedWeb · web corpus · data filtering
Official artifacts: Dataset
Metadata checked 2026-07-28
2023ICCVPeer reviewed
TinyCLIP: CLIP Distillation via Affinity Mimicking and Weight Inheritance
Kan Wu, Houwen Peng, Zhenghong Zhou, Bin Xiao, Mengchen Liu, Lu Yuan, Hong Xuan, Michael Valenzuela, Xi Chen, Xinggang Wang, Hongyang Chao, and Han Hu
L2–L3
Knowledge distillation Image–text representation Multimodal retrieval
knowledge distillation / model transfer · image–text representation · multimodal embedding / retrieval · CLIP compression · affinity mimicking · weight inheritance · vision-language retrieval
Official artifacts: Official code and checkpoints
Metadata checked 2026-08-10
2023NeurIPSPeer reviewed
Toolformer: Language Models Can Teach Themselves to Use Tools
Timo Schick et al.
L2–L3
RAG / tools / agents
RAG / tools / agents · Toolformer · tool use · self-supervision
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
Tree of Thoughts: Deliberate Problem Solving with Large Language Models
Shunyu Yao et al.
L2–L3
Reasoning / verification
LLM reasoning / verification · Tree of Thoughts · search · deliberate reasoning
Official artifacts: Official code
Metadata checked 2026-08-08
2023EMNLP System DemonstrationsPeer reviewed
Video-LLaMA: An Instruction-tuned Audio-Visual Language Model for Video Understanding
Hang Zhang, Xin Li, and Lidong Bing
L2
Generation / video / agents
multimodal generation / agents · video MLLM · audio-visual · instruction tuning
Metadata checked 2026-07-27
2023NeurIPSPeer reviewed
Visual Instruction Tuning
Haotian Liu et al.
L2–L3
Multimodal LLMs
multimodal LLM · LLaVA · visual instruction tuning · pathology VLM
Metadata checked 2026-07-27
2023ICLRPeer reviewed
When and Why Vision-Language Models Behave like Bags-of-Words, and What to Do About It?
Mert Yuksekgonul et al.
L2
V&L evaluation / safety
vision–language evaluation / safety · ARO · compositionality · word order
Metadata checked 2026-07-28
2022CVPRPeer reviewed
A ConvNet for the 2020s
Zhuang Liu et al.
L2
Image / recognition Visual representations
image recognition · visual representation learning · ConvNeXt · CNN · modern training
Metadata checked 2026-07-27
2022ECCVPeer reviewed
A-OKVQA: A Benchmark for Visual Question Answering Using World Knowledge
Dustin Schwenk et al.
L2
Captioning / VQA V&L evaluation / safety
captioning / VQA · vision–language evaluation / safety · knowledge VQA · rationales · world knowledge
Metadata checked 2026-07-28
2022ICLRPeer reviewed
BEiT: BERT Pre-Training of Image Transformers
Hangbo Bao et al.
Research
Visual representations
self-supervised learning · BEiT · masked modeling · visual representation learning
Metadata checked 2026-07-27
2022ICMLPeer reviewed
BLIP: Bootstrapping Language-Image Pre-training for Unified Vision-Language Understanding and Generation
Junnan Li et al.
L2
Image–text representation Captioning / VQA
image–text representation · captioning / VQA · contrastive learning · image-text matching · caption filtering
Metadata checked 2026-07-27
2022TACLPeer reviewed
ByT5: Towards a Token-Free Future with Pre-trained Byte-to-Byte Models
Linting Xue, Aditya Barua, Noah Constant, Rami Al-Rfou, Sharan Narang, Mihir Kale, Adam Roberts, and Colin Raffel
L2
Text / sequence models Transformers / pretraining Multilingual / low resource
text / sequence modeling · NLP transformers / pretraining · ByT5 · byte-level modeling · tokenization · multilingual / low-resource NLP
Official artifacts: Code and checkpoints
Metadata checked 2026-07-28
2022ECCVPeer reviewed
ByteTrack: Multi-Object Tracking by Associating Every Detection Box
Yifu Zhang et al.
L2
Video / 3D / embodied
video / 3D / embodied vision · multi-object tracking · data association · low-score detections
Official artifacts: Code
Metadata checked 2026-07-28
2022NeurIPSPeer reviewed
Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
Jason Wei et al.
L2
LLM post-training Reasoning / verification
LLM post-training / alignment · chain of thought · reasoning · prompting · LLM reasoning / verification
Metadata checked 2026-07-27
2022Findings of ACLPeer reviewed
ChartQA: A Benchmark for Question Answering about Charts with Visual and Logical Reasoning
Ahmed Masry et al.
L2
Captioning / VQA V&L evaluation / safety Document intelligence
captioning / VQA · vision–language evaluation / safety · chart question answering · arithmetic reasoning · visual attributes · document intelligence
Metadata checked 2026-07-28
2022arXivPreprint
Classifier-Free Diffusion Guidance
Jonathan Ho and Tim Salimans
L2–L3
Generative / representation
generative / representation learning · classifier-free guidance · conditional diffusion · guidance
Metadata checked 2026-07-27
2022TMLRPeer reviewed
CoCa: Contrastive Captioners are Image-Text Foundation Models
Jiahui Yu et al.
L2
Image–text representation Captioning / VQA Loss / objectives
pathology VLM · CoCa · contrastive captioning · image–text pretraining · image–text representation · captioning / VQA · loss / objective design
Metadata checked 2026-07-27
2022SciencePeer reviewed
Competition-Level Code Generation with AlphaCode
Yujia Li, David Choi, Junyoung Chung, Nate Kushman, Julian Schrittwieser, Rémi Leblond, Tom Eccles, James Keeling, Felix Gimeno, Agustin Dal Lago, Thomas Hubert, Peter Choy, Cyprien de Masson d’Autume, Igor Babuschkin, Xinyun Chen, Po-Sen Huang, Johannes Welbl, Sven Gowal, Alexey Cherepanov, James Molloy, Daniel J. Mankowitz, Esme Sutherland Robson, Pushmeet Kohli, Nando de Freitas, Koray Kavukcuoglu, and Oriol Vinyals
L2–L3
Code / software engineering
code generation / software engineering · AlphaCode · competitive programming · large-scale sampling · behavioral clustering · CodeContests
Official artifacts: Author manuscript · Official CodeContests dataset
Metadata checked 2026-08-20
2022arXivPreprint
Constitutional AI: Harmlessness from AI Feedback
Yuntao Bai et al.
L2–L3
LLM post-training LLM evaluation / safety
LLM post-training / alignment · LLM evaluation / safety · constitutional AI · AI feedback · safety
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Decoupled Knowledge Distillation
Borui Zhao, Quan Cui, Renjie Song, Yiyu Qiu, and Jiajun Liang
L2–L3
Knowledge distillation Visual representations Detection / segmentation
knowledge distillation / model transfer · visual representation learning · detection / segmentation · target-class knowledge · non-target-class knowledge · logit distillation · dark knowledge
Official artifacts: Official code
Metadata checked 2026-08-10
2022ACLPeer reviewed
Deduplicating Training Data Makes Language Models Better
Katherine Lee et al.
L2
LLM pretraining
LLM pretraining / scaling · deduplication · pretraining data · memorization
Official artifacts: Code
Metadata checked 2026-07-28
2022NeurIPSPeer reviewed
Denoising Diffusion Restoration Models
Bahjat Kawar, Michael Elad, Stefano Ermon, and Jiaming Song
L2–L3
Restoration / inverse problems
DDRM · diffusion posterior sampling · linear inverse problems · image restoration
Official artifacts: Official implementation · arXiv record
Metadata checked 2026-08-20
2022ECCVPeer reviewed
Detecting Twenty-thousand Classes using Image-level Supervision
Xingyi Zhou et al.
L2
Grounding / open vocabulary
grounding / open vocabulary · Detic · open vocabulary · image-level supervision
Metadata checked 2026-07-27
2022CVPRPeer reviewed
DTFD-MIL: Double-Tier Feature Distillation Multiple Instance Learning for Histopathology Whole Slide Image Classification
Hongrun Zhang et al.
L2–L3
Knowledge distillation
MIL / WSI · DTFD-MIL · pseudo-bags · feature distillation · knowledge distillation / model transfer
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
Elucidating the Design Space of Diffusion-Based Generative Models
Tero Karras, Miika Aittala, Timo Aila, and Samuli Laine
L2–L3
Generative / representation Generative vision Generation / video / agents Restoration / inverse problems
generative / representation learning · generative vision · diffusion models · preconditioning · sampling design · multimodal generation / agents
Official artifacts: Code
Metadata checked 2026-07-28
2022ICLRPeer reviewed
FILIP: Fine-grained Interactive Language-Image Pre-Training
Lewei Yao et al.
L2
Image–text representation Loss / objectives
image–text representation · late interaction · token alignment · contrastive learning · loss / objective design
Metadata checked 2026-07-27
2022ICLRPeer reviewed
Finetuned Language Models Are Zero-Shot Learners
Jason Wei et al.
L2
LLM post-training
LLM post-training / alignment · FLAN · instruction tuning · zero-shot transfer
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
Flamingo: a Visual Language Model for Few-Shot Learning
Jean-Baptiste Alayrac et al.
L2
Multimodal LLMs
multimodal LLM · Perceiver resampler · gated cross-attention · few-shot
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
FlashAttention: Fast and Memory-Efficient Exact Attention with IO-Awareness
Tri Dao et al.
L2–L3
Transformers / pretraining LLM pretraining Neural networks LLM systems
NLP transformers / pretraining · LLM pretraining / scaling · neural networks · FlashAttention · IO awareness · efficient attention · LLM training / serving systems
Metadata checked 2026-07-27
2022CVPRPeer reviewed
FLAVA: A Foundational Language And Vision Alignment Model
Amanpreet Singh et al.
L2
Image–text representation
image–text representation · multimodal pretraining · unimodal objectives · fusion
Metadata checked 2026-07-27
2022ECCVPeer reviewed
FOSTER: Feature Boosting and Compression for Class-Incremental Learning
Fu-Yun Wang, Da-Wei Zhou, Han-Jia Ye, and De-Chuan Zhan
L3
Published in the ECCV 2022 main conference proceedings, pages 398–414.
Knowledge distillation Transfer / adaptation / continual Image / recognition
knowledge distillation / model transfer · image recognition · FOSTER · class-incremental learning · feature boosting · residual correction · balanced distillation
Official artifacts: Official implementation
Metadata checked 2026-09-02
2022TMLRPeer reviewed
GIT: A Generative Image-to-text Transformer for Vision and Language
Jianfeng Wang et al.
L2
Captioning / VQA
captioning / VQA · generative modeling · captioning · VQA
Metadata checked 2026-07-27
2022ICMLPeer reviewed
GLIDE: Towards Photorealistic Image Generation and Editing with Text-Guided Diffusion Models
Alexander Quinn Nichol et al.
L2
Generation / video / agents
multimodal generation / agents · text-to-image diffusion · classifier-free guidance · image editing
Metadata checked 2026-07-28
2022CVPRPeer reviewed
Grounded Language-Image Pre-training
Liunian Harold Li et al.
L3
Detection / segmentation Grounding / open vocabulary
detection / segmentation · GLIP · phrase grounding · open-vocabulary detection · grounding / open vocabulary
Metadata checked 2026-07-27
2022CVPRPeer reviewed
GroupViT: Semantic Segmentation Emerges from Text Supervision
Jiarui Xu et al.
L2
Image–text representation Grounding / open vocabulary
image–text representation · grounding / open vocabulary · semantic grouping · text supervision · zero-shot segmentation
Metadata checked 2026-07-28
2022CVPRPeer reviewed
High-Resolution Image Synthesis with Latent Diffusion Models
Robin Rombach et al.
L2–L3
Generative / representation Generative vision Generation / video / agents Restoration / inverse problems
generative / representation learning · latent diffusion · text-to-image · autoencoder · generative vision · multimodal generation / agents · generative / virtual pathology
Metadata checked 2026-07-27
2022ICLRPeer reviewed
iBOT: Image BERT Pre-Training with Online Tokenizer
Jinghao Zhou et al.
Research
Visual representations
self-supervised learning · iBOT · masked modeling · self-distillation · visual representation learning
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Image Segmentation Using Text and Image Prompts
Timo Lüddecke and Alexander S. Ecker
L2
Grounding / open vocabulary
grounding / open vocabulary · CLIPSeg · text prompts · image prompts
Metadata checked 2026-07-28
2022ICMLPeer reviewed
Improving Language Models by Retrieving from Trillions of Tokens
Sebastian Borgeaud et al.
L3
RAG / tools / agents
RAG / tools / agents · RETRO · retrieval · language-model pretraining
Metadata checked 2026-07-27
2022arXivTechnical report
InternVideo: General Video Foundation Models via Generative and Discriminative Learning
Yi Wang et al.
L2
Generation / video / agents
multimodal generation / agents · video foundation model · masked modeling · video-text contrast
Metadata checked 2026-07-27
2022NAACLPeer reviewed
KAT: A Knowledge Augmented Transformer for Vision-and-Language
Liangke Gui et al.
L2
Captioning / VQA
captioning / VQA · knowledge retrieval · VQA · transformer
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Knowledge Distillation: A Good Teacher Is Patient and Consistent
Lucas Beyer, Xiaohua Zhai, Amelie Royer, Larisa Markeeva, Rohan Anil, and Alexander Kolesnikov
L2–L3
Knowledge distillation Visual representations Evaluation / reproducibility
knowledge distillation / model transfer · visual representation learning · evaluation / reproducibility · function matching · consistent augmentation · long training · image classification
Official artifacts: Teacher model code and weights
Metadata checked 2026-08-10
2022NeurIPS Datasets and BenchmarksPeer reviewed
LAION-5B: An Open Large-Scale Dataset for Training Next Generation Image-Text Models
Christoph Schuhmann et al.
L2–L3
Image–text representation V&L evaluation / safety
image–text representation · vision–language evaluation / safety · image-text data · web scale · data provenance
Metadata checked 2026-07-28
2022ACM MultimediaPeer reviewed
LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking
Yupan Huang, Tengchao Lv, Lei Cui, Yutong Lu, and Furu Wei
L2–L3
Document intelligence Image–text representation
document intelligence · image–text representation · LayoutLMv3 · unified text-image masking · word-patch alignment
Official artifacts: Official implementation and models
Metadata checked 2026-08-20
2022CVPRPeer reviewed
LiT: Zero-Shot Transfer with Locked-image Text Tuning
Xiaohua Zhai et al.
L2
Image–text representation Multimodal retrieval
image–text representation · locked image encoder · dual encoder · zero-shot · multimodal embedding / retrieval
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
Locating and Editing Factual Associations in GPT
Kevin Meng, David Bau, Alex Andonian, and Yonatan Belinkov
L2–L3
LLM interpretability
LLM interpretability / editing / unlearning · ROME · causal tracing · model editing · factual associations
Official artifacts: Official project and code
Metadata checked 2026-08-08
2022ICLRPeer reviewed
LoRA: Low-Rank Adaptation of Large Language Models
Edward J. Hu et al.
L2
Transformers / pretraining LLM post-training Generative / representation Transfer / adaptation / continual
NLP transformers / pretraining · LLM post-training / alignment · generative / representation learning · LoRA · parameter-efficient tuning · low rank
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Masked Autoencoders Are Scalable Vision Learners
Kaiming He et al.
L2
Computer vision Generative / representation Visual representations
computer vision · self-supervised learning · MAE · masked modeling · generative / representation learning · visual representation learning
Metadata checked 2026-07-27
2022ECCVPeer reviewed
Masked Generative Distillation
Zhendong Yang, Zhe Li, Mingqi Shao, Dachuan Shi, Zehuan Yuan, and Chun Yuan
L2–L3
Knowledge distillation Visual representations Detection / segmentation
knowledge distillation / model transfer · visual representation learning · detection / segmentation · masked feature modeling · generative distillation · dense prediction · feature reconstruction
Official artifacts: Official code
Metadata checked 2026-08-10
2022CVPRPeer reviewed
Masked-Attention Mask Transformer for Universal Image Segmentation
Bowen Cheng et al.
L2–L3
Detection / segmentation
detection / segmentation · Mask2Former · universal segmentation · masked attention
Metadata checked 2026-07-27
2022ICMLPeer reviewed
Multi-Grained Vision Language Pre-Training: Aligning Texts with Visual Concepts
Yan Zeng, Xinsong Zhang, and Hang Li
L2–L3
Image–text representation Grounding / open vocabulary
image–text representation · grounding / open vocabulary · X-VLM · multi-grained alignment · visual grounding · vision-language pretraining
Official artifacts: Official implementation and checkpoints
Metadata checked 2026-08-20
2022ECCVPeer reviewed
OCR-Free Document Understanding Transformer
Geewook Kim et al.
L2–L3
Document intelligence Captioning / VQA
document intelligence · captioning / VQA · Donut · OCR-free document understanding · structured decoding
Official artifacts: Official code
Metadata checked 2026-08-08
2022ICMLPeer reviewed
OFA: Unifying Architectures, Tasks, and Modalities Through a Simple Sequence-to-Sequence Learning Framework
Peng Wang et al.
L2
Captioning / VQA
captioning / VQA · sequence-to-sequence · task unification · multimodal
Metadata checked 2026-07-27
2022ICLRPeer reviewed
Offline Reinforcement Learning with Implicit Q-Learning
Ilya Kostrikov, Ashvin Nair, and Sergey Levine
L2–L3
AI / reinforcement learning
AI / reinforcement learning · implicit Q-learning · offline RL · distribution shift
Metadata checked 2026-07-28
2022NeurIPSPeer reviewed
Photorealistic Text-to-Image Diffusion Models with Deep Language Understanding
Chitwan Saharia et al.
L2
Generation / video / agents
multimodal generation / agents · Imagen · text-to-image · cascaded diffusion
Metadata checked 2026-07-27
2022ICLRPeer reviewed
PolyLoss: A Polynomial Expansion Perspective of Classification Loss Functions
Zhaoqi Leng, Mingxing Tan, Chenxi Liu, Ekin Dogus Cubuk, Xiaojie Shi, Shuyang Cheng, and Dragomir Anguelov
L2–L3
Loss / objectives Image / recognition Detection / segmentation
loss / objective design · image recognition · detection / segmentation · polynomial loss · Poly-1 · cross entropy · focal loss · classification objective
Official artifacts: Author-linked EfficientNetV2 implementation
Metadata checked 2026-08-10
2022ICLRPeer reviewed
Progressive Distillation for Fast Sampling of Diffusion Models
Tim Salimans and Jonathan Ho
L2–L3
Knowledge distillation Generative / representation Generative vision
knowledge distillation / model transfer · generative / representation learning · generative vision · progressive distillation · diffusion sampling · few-step generation
Official artifacts: Official code
Metadata checked 2026-08-10
2022CVPRPeer reviewed
PubTables-1M: Towards Comprehensive Table Extraction From Unstructured Documents
Brandon Smock, Rohith Pesala, and Robin Abraham
L2–L3
Document intelligence
document intelligence · PubTables-1M · table extraction · table structure recognition · dataset canonicalization
Official artifacts: Official dataset, code, and Table Transformer models
Metadata checked 2026-08-20
2022CVPRPeer reviewed
RegionCLIP: Region-Based Language-Image Pretraining
Yiwu Zhong et al.
L2
Grounding / open vocabulary
grounding / open vocabulary · region-text alignment · open-vocabulary detection
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Restormer: Efficient Transformer for High-Resolution Image Restoration
Syed Waqas Zamir et al.
L2–L3
Generative vision Restoration / inverse problems
generative vision · Restormer · image restoration · efficient attention
Metadata checked 2026-07-27
2022ECCVPeer reviewed
Simple Baselines for Image Restoration
Liangyu Chen, Xiaojie Chu, Xiangyu Zhang, and Jian Sun
L3
Published in the ECCV 2022 proceedings; ECVA hosts the canonical open proceedings PDF.
Restoration / inverse problems
NAFNet · image restoration · SimpleGate · denoising · deblurring
Official artifacts: Official implementation and models
Metadata checked 2026-08-30
2022ECCVPeer reviewed
Simple Open-Vocabulary Object Detection with Vision Transformers
Matthias Minderer et al.
L2
Grounding / open vocabulary
grounding / open vocabulary · OWL-ViT · open-vocabulary detection · vision transformer
Metadata checked 2026-07-27
2022ECCVPeer reviewed
SLIP: Self-supervision meets Language-Image Pre-training
Norman Mu et al.
L2
Image–text representation
image–text representation · self-supervision · language-image pretraining · contrastive learning
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
STaR: Bootstrapping Reasoning With Reasoning
Eric Zelikman, Yuhuai Wu, Jesse Mu, and Noah D. Goodman
L2–L3
Reasoning / verification
LLM reasoning / verification · STaR · rationale bootstrapping · self-improvement
Official artifacts: Official code
Metadata checked 2026-08-08
2022JMLRPeer reviewed
Switch Transformers: Scaling to Trillion Parameter Models with Simple and Efficient Sparsity
William Fedus, Barret Zoph, and Noam Shazeer
L2–L3
LLM pretraining
LLM pretraining / scaling · mixture of experts · sparse model · scaling
Metadata checked 2026-07-27
2022TACLPeer reviewed
The FLORES-101 Evaluation Benchmark for Low-Resource and Multilingual Machine Translation
Naman Goyal, Cynthia Gao, Vishrav Chaudhary, Peng-Jen Chen, Guillaume Wenzek, Da Ju, Sanjana Krishnan, Marc'Aurelio Ranzato, Francisco Guzmán, and Angela Fan
L2–L3
Multilingual / low resource Generation / MT evaluation
multilingual / low-resource NLP · generation / translation evaluation · FLORES-101 · multilingual translation benchmark · low-resource languages · spBLEU
Official artifacts: Official FLORES repository
Metadata checked 2026-08-10
2022ICLRPeer reviewed
Train Short, Test Long: Attention with Linear Biases Enables Input Length Extrapolation
Ofir Press, Noah A. Smith, and Mike Lewis
L2
Transformers / pretraining LLM pretraining
NLP transformers / pretraining · LLM pretraining / scaling · ALiBi · position bias · length extrapolation
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
Training Compute-Optimal Large Language Models
Jordan Hoffmann et al.
L2
LLMs LLM pretraining
LLM · scaling laws · compute · LLM pretraining / scaling
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
Training Language Models to Follow Instructions with Human Feedback
Long Ouyang et al.
L2
LLMs LLM post-training
LLM · instruction tuning · RLHF · LLM post-training / alignment
Metadata checked 2026-07-27
2022ACLPeer reviewed
TruthfulQA: Measuring How Models Mimic Human Falsehoods
Stephanie Lin, Jacob Hilton, and Owain Evans
L2
LLM evaluation / safety
LLM evaluation / safety · TruthfulQA · truthfulness · misconceptions
Metadata checked 2026-07-27
2022ICLRPeer reviewed
VICReg: Variance-Invariance-Covariance Regularization for Self-Supervised Learning
Adrien Bardes, Jean Ponce, and Yann LeCun
Research
Generative / representation Visual representations
self-supervised learning · VICReg · collapse avoidance · generative / representation learning · visual representation learning
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Video Swin Transformer
Ze Liu et al.
L2
Video / 3D / embodied
video / 3D / embodied vision · Video Swin · video transformer · local attention
Metadata checked 2026-07-27
2022NeurIPSPeer reviewed
VideoMAE: Masked Autoencoders Are Data-Efficient Learners for Self-Supervised Video Pre-Training
Zhan Tong et al.
L2–L3
Video / 3D / embodied
video / 3D / embodied vision · VideoMAE · masked modeling · video SSL
Metadata checked 2026-07-27
2022CVPRPeer reviewed
Winoground: Probing Vision and Language Models for Visio-Linguistic Compositionality
Tristan Thrush et al.
L2
V&L evaluation / safety
vision–language evaluation / safety · compositionality · minimal pairs · image-text matching
Metadata checked 2026-07-27
2021NAACLPeer reviewed
A Frustratingly Easy Approach for Entity and Relation Extraction
Zexuan Zhong and Danqi Chen
L2–L3
Structure / extraction
linguistic structure / information extraction · entity extraction · relation extraction · pipeline modeling · typed entity markers
Official artifacts: Official implementation
Metadata checked 2026-08-10
2021Transformer Circuits ThreadTechnical report
A Mathematical Framework for Transformer Circuits
Nelson Elhage et al.
L2–L3
Author-published technical article; not presented as a peer-reviewed conference paper.
LLM interpretability Interpretability
LLM interpretability / editing / unlearning · transformer circuits · attention composition · mechanistic interpretability · interpretability / attribution
Metadata checked 2026-08-08
2021NeurIPSPeer reviewed
Align Before Fuse: Vision and Language Representation Learning with Momentum Distillation
Junnan Li et al.
L2
Image–text representation Knowledge distillation
image–text representation · alignment · fusion · momentum distillation · hard negatives · knowledge distillation / model transfer
Metadata checked 2026-07-27
2021ICLRPeer reviewed
An Image is Worth 16×16 Words: Transformers for Image Recognition at Scale
Alexey Dosovitskiy et al.
L2
Computer vision Visual representations Image / recognition
computer vision · vision transformer · visual representation learning · image recognition
Metadata checked 2026-07-27
2021ICMLPeer reviewed
Barlow Twins: Self-Supervised Learning via Redundancy Reduction
Jure Zbontar et al.
Research
Generative / representation Visual representations
self-supervised learning · Barlow Twins · redundancy reduction · generative / representation learning · visual representation learning
Metadata checked 2026-07-27
2021NeurIPS Datasets and BenchmarksPeer reviewed
BEIR: A Heterogeneous Benchmark for Zero-shot Evaluation of Information Retrieval Models
Nandan Thakur et al.
L2–L3
NLP tasks / evaluation
NLP tasks / evaluation · BEIR · zero-shot retrieval · distribution shift
Official artifacts: Benchmark code and data
Metadata checked 2026-07-28
2021ICCVPeer reviewed
Channel-Wise Knowledge Distillation for Dense Prediction
Changyong Shu, Yifan Liu, Jianfei Gao, Zheng Yan, and Chunhua Shen
L3
ICCV 2021 proceedings paper and author v4 manuscript checked; older missing prerequisite added during the October 2026 audit, not a new 2026 paper. Official repository records the semantic-segmentation code release on 20 August 2021.
Knowledge distillation Loss / objectives Detection / segmentation
knowledge distillation / model transfer · loss / objective design · detection / segmentation · knowledge-distillation · fitnets
Official artifacts: Official implementation · Author manuscript
Metadata checked 2026-10-02
2021Communications of the ACMPeer reviewed
Datasheets for Datasets
Timnit Gebru et al.
L1–L2
Evaluation / reproducibility NLP tasks / evaluation LLM evaluation / safety Vision evaluation
evaluation / reproducibility · datasheets · dataset documentation · data provenance · NLP tasks / evaluation · LLM evaluation / safety · vision evaluation / robustness
Metadata checked 2026-07-27
2021NeurIPSPeer reviewed
Decision Transformer: Reinforcement Learning via Sequence Modeling
Lili Chen et al.
L3
AI / reinforcement learning
AI / reinforcement learning · Decision Transformer · offline RL · sequence modeling
Metadata checked 2026-07-27
2021ICLRPeer reviewed
Deformable DETR: Deformable Transformers for End-to-End Object Detection
Xizhou Zhu et al.
L2–L3
Detection / segmentation
detection / segmentation · Deformable DETR · set prediction · multiscale attention
Metadata checked 2026-07-27
2021ICLRPeer reviewed
Denoising Diffusion Implicit Models
Jiaming Song, Chenlin Meng, and Stefano Ermon
L2–L3
Generative / representation Restoration / inverse problems
generative / representation learning · DDIM · diffusion sampling · implicit model
Metadata checked 2026-07-27
2021CVPRPeer reviewed
Distilling Knowledge via Knowledge Review
Pengguang Chen, Shu Liu, Hengshuang Zhao, and Jiaya Jia
L2–L3
Knowledge distillation Visual representations Detection / segmentation
knowledge distillation / model transfer · visual representation learning · detection / segmentation · knowledge review · cross-stage connection · attention-based fusion · hierarchical context loss
Official artifacts: Official code
Metadata checked 2026-08-10
2021NeurIPSPeer reviewed
Do Transformers Really Perform Bad for Graph Representation?
Chengxuan Ying, Tianle Cai, Shengjie Luo, Shuxin Zheng, Guolin Ke, Di He, Yanming Shen, and Tie-Yan Liu
L3
Published in the NeurIPS 2021 proceedings.
Graph / geometric ML Neural networks
graph / geometric learning · neural networks · Graphormer · graph transformer · centrality encoding · spatial encoding · edge encoding
Official artifacts: Official implementation
Metadata checked 2026-08-30
2021ICCVPeer reviewed
DocFormer: End-to-End Transformer for Document Understanding
Srikar Appalaraju, Bhavan Jasani, Bhargava Urala Kota, Yusheng Xie, and R. Manmatha
L2–L3
Document intelligence
document intelligence · DocFormer · multimodal self-attention · spatial embeddings
Metadata checked 2026-08-20
2021WACVPeer reviewed
DocVQA: A Dataset for VQA on Document Images
Minesh Mathew et al.
L2
Captioning / VQA V&L evaluation / safety Document intelligence
captioning / VQA · vision–language evaluation / safety · document VQA · OCR · layout reasoning · document intelligence
Metadata checked 2026-07-28
2021ICMLPeer reviewed
E(n) Equivariant Graph Neural Networks
Víctor Garcia Satorras, Emiel Hoogeboom, and Max Welling
L2–L3
Graph / geometric ML Neural networks
graph / geometric learning · neural networks · E(n) equivariance · geometric deep learning · message passing
Official artifacts: Official code
Metadata checked 2026-08-08
2021SCPeer reviewed
Efficient Large-Scale Language Model Training on GPU Clusters Using Megatron-LM
Deepak Narayanan et al.
L2–L3
LLM systems LLM pretraining
LLM training / serving systems · LLM pretraining / scaling · Megatron-LM · tensor parallelism · pipeline parallelism
Official artifacts: Megatron-LM
Metadata checked 2026-08-08
2021ICCVPeer reviewed
Emerging Properties in Self-Supervised Vision Transformers
Mathilde Caron et al.
L2
Computer vision Generative / representation Visual representations Knowledge distillation
computer vision · self-supervised learning · DINO · self-distillation · vision transformer · generative / representation learning · visual representation learning · knowledge distillation / model transfer
Metadata checked 2026-07-27
2021arXivPreprint
Evaluating Large Language Models Trained on Code
Mark Chen et al.
L2–L3
Code / software engineering
code generation / software engineering · Codex · HumanEval · functional correctness · pass@k · code generation
Official artifacts: Official HumanEval harness
Metadata checked 2026-08-20
2021CVPRPeer reviewed
Exploring Simple Siamese Representation Learning
Xinlei Chen and Kaiming He
Research
Visual representations
self-supervised learning · SimSiam · stop-gradient · visual representation learning
Metadata checked 2026-07-27
2021USENIX SecurityPeer reviewed
Extracting Training Data from Large Language Models
Nicholas Carlini et al.
L2–L3
LLM evaluation / safety
LLM evaluation / safety · training-data extraction · privacy · memorization
Official artifacts: Code
Metadata checked 2026-07-28
2021ICCVPeer reviewed
Frozen in Time: A Joint Video and Image Encoder for End-to-End Retrieval
Max Bain et al.
L2
Generation / video / agents
multimodal generation / agents · video-text retrieval · joint image-video encoder
Metadata checked 2026-07-27
2021JMLRPeer reviewed
Improving Reproducibility in Machine Learning Research
Joelle Pineau et al.
L1–L2
Evaluation / reproducibility
evaluation / reproducibility · reproducibility · checklist · code submission
Metadata checked 2026-07-27
2021ICMLPeer reviewed
Is Space-Time Attention All You Need for Video Understanding?
Gedas Bertasius, Heng Wang, and Lorenzo Torresani
L2
Video / 3D / embodied
video / 3D / embodied vision · TimeSformer · video transformer · space-time attention
Metadata checked 2026-07-27
2021CVPRPeer reviewed
KRISP: Integrating Implicit and Symbolic Knowledge for Open-Domain Knowledge-Based VQA
Kenneth Marino et al.
L2
Captioning / VQA
captioning / VQA · knowledge graph · VQA · symbolic reasoning
Metadata checked 2026-07-27
2021ACL-IJCNLPPeer reviewed
LayoutLMv2: Multi-modal Pre-training for Visually-rich Document Understanding
Yang Xu, Yiheng Xu, Tengchao Lv, Lei Cui, Furu Wei, Guoxin Wang, Yijuan Lu, Dinei Florencio, Cha Zhang, Wanxiang Che, Min Zhang, and Lidong Zhou
L2–L3
Document intelligence Image–text representation
document intelligence · image–text representation · layout-aware pretraining · visual-language pretraining · OCR
Official artifacts: Official implementation
Metadata checked 2026-08-10
2021ICMLPeer reviewed
Learning Transferable Visual Models From Natural Language Supervision
Alec Radford et al.
L2
Image–text representation Detection / segmentation Visual representations Multimodal retrieval Loss / objectives
image–text representation · pathology VLM · CLIP · image–text alignment · zero-shot · detection / segmentation · visual representation learning · multimodal embedding / retrieval · loss / objective design
Metadata checked 2026-07-27
2021EACLPeer reviewed
Leveraging Passage Retrieval with Generative Models for Open Domain Question Answering
Gautier Izacard and Edouard Grave
L2
First released as a preprint in 2020; the catalog records the peer-reviewed EACL 2021 version of record.
RAG / tools / agents
RAG / tools / agents · Fusion-in-Decoder · retrieval · question answering
Metadata checked 2026-07-28
2021ICLRPeer reviewed
Long-Tail Learning via Logit Adjustment
Aditya Krishna Menon, Sadeep Jayasumana, Ankit Singh Rawat, Himanshu Jain, Andreas Veit, and Sanjiv Kumar
L2–L3
Loss / objectives Vision evaluation
loss / objective design · vision evaluation / robustness · logit adjustment · long-tail learning · class prior
Metadata checked 2026-08-10
2021ICCVPeer reviewed
MDETR: Modulated Detection for End-to-End Multi-Modal Understanding
Aishwarya Kamath et al.
L2
Grounding / open vocabulary
grounding / open vocabulary · phrase grounding · language-conditioned detection · DETR
Metadata checked 2026-07-27
2021NeurIPS Datasets and BenchmarksPeer reviewed
Measuring Coding Challenge Competence With APPS
Dan Hendrycks, Steven Basart, Saurav Kadavath, Mantas Mazeika, Akul Arora, Ethan Guo, Collin Burns, Samir Puranik, Horace He, Dawn Song, and Jacob Steinhardt
L2–L3
Code / software engineering
code generation / software engineering · APPS · code-generation benchmark · execution-based evaluation · programming difficulty · strict accuracy
Official artifacts: Official implementation, evaluation harness, and data access · Author manuscript
Metadata checked 2026-08-20
2021ICLRPeer reviewed
Measuring Massive Multitask Language Understanding
Dan Hendrycks et al.
L1–L2
LLM evaluation / safety
LLM evaluation / safety · MMLU · knowledge evaluation · benchmark
Metadata checked 2026-07-27
2021IJCAIPeer reviewed
MRD-Net: Multi-Modal Residual Knowledge Distillation for Spoken Question Answering
Chenyu You, Nuo Chen, and Yuexian Zou
L3
Published in the IJCAI 2021 main track, pages 3985–3991, DOI 10.24963/ijcai.2021/549.
Knowledge distillation Text / sequence models Transformers / pretraining NLP tasks / evaluation
knowledge distillation / model transfer · text / sequence modeling · NLP transformers / pretraining · NLP tasks / evaluation · MRD-Net · spoken question answering · audio assistant · ASR errors · multimodal distillation
Metadata checked 2026-09-02
2021NAACL-HLTPeer reviewed
mT5: A Massively Multilingual Pre-trained Text-to-Text Transformer
Linting Xue et al.
L2
Transformers / pretraining Multilingual / low resource
NLP transformers / pretraining · mT5 · multilingual generation · text-to-text transfer · multilingual / low-resource NLP
Official artifacts: Code and checkpoints
Metadata checked 2026-07-28
2021CVPRPeer reviewed
Multi-Scale Aligned Distillation for Low-Resolution Detection
Lu Qi, Jason Kuen, Jiuxiang Gu, Zhe Lin, Yi Wang, Yukang Chen, Yanwei Li, and Jiaya Jia
L2–L3
Peer-reviewed CVPR 2021 proceedings paper.
Knowledge distillation Detection / segmentation Vision evaluation
knowledge distillation / model transfer · detection / segmentation · vision evaluation / robustness · MSAD · low-resolution detection · multiscale alignment · feature distillation · scale mismatch
Official artifacts: Official code
Metadata checked 2026-08-23
2021NeurIPSPeer reviewed
Multimodal Few-Shot Learning with Frozen Language Models
Maria Tsimpoukelli et al.
L2
The peer-reviewed version of record was published at NeurIPS 2021.
Multimodal LLMs
multimodal LLM · frozen language model · visual prefix · few-shot learning
Metadata checked 2026-07-28
2021ICCVPeer reviewed
Pyramid Vision Transformer: A Versatile Backbone for Dense Prediction without Convolutions
Wenhai Wang et al.
L2
Visual representations
visual representation learning · PVT · vision transformer · dense prediction
Metadata checked 2026-07-27
2021NeurocomputingPeer reviewed
Residual error based knowledge distillation
Mengya Gao, Yujun Wang, and Liang Wan
L3
Published in Neurocomputing 433 (14 April 2021), pages 154–161, DOI 10.1016/j.neucom.2020.10.113. The differently authored arXiv:2002.09168 is recorded only as a related precursor, not as this journal version.
Knowledge distillation Image / recognition Visual representations
knowledge distillation / model transfer · image recognition · visual representation learning · ResErrKD · residual error · assistant network · feature distillation · capacity gap
Metadata checked 2026-09-02
2021IEEE Transactions on Image ProcessingPeer reviewed
ResKD: Residual-Guided Knowledge Distillation
Xuewei Li, Songyuan Li, Bourahla Omar, Fei Wu, and Xi Li
L3
Published in IEEE Transactions on Image Processing 30 (2021), pages 4735–4746.
Knowledge distillation Image / recognition Vision evaluation
knowledge distillation / model transfer · image recognition · vision evaluation / robustness · ResKD · residual logits · res-student · sample-adaptive inference · neural architecture search
Official artifacts: Author manuscript
Metadata checked 2026-09-02
2021IEEE Transactions on Image ProcessingPeer reviewed
Resolution-Aware Knowledge Distillation for Efficient Inference
Zhanxiang Feng, Jianhuang Lai, and Xiaohua Xie
L3
Peer-reviewed journal article, volume 30, pages 6985–6996. The paper calls its framework RKD, meaning resolution-aware—not relational—knowledge distillation. Teacher and student use the same architecture but paired high- and low-resolution inputs; only the low-resolution student is needed at inference.
Knowledge distillation Image / recognition Visual representations Detection / segmentation Vision evaluation
knowledge distillation / model transfer · image recognition · visual representation learning · detection / segmentation · vision evaluation / robustness · resolution-aware knowledge distillation · cross-resolution distillation · low-resolution inference · adversarial feature alignment · maximum mean discrepancy
Official artifacts: PubMed record
Metadata checked 2026-09-02
2021Artificial Intelligence in MedicinePeer reviewed
Resolution-Based Distillation for Efficient Histology Image Classification
Joseph DiPalma, Arief A. Suriawinata, Laura J. Tafe, Lorenzo Torresani, and Saeed Hassanpour
L3
Peer-reviewed journal article 119:102136.
Knowledge distillation
knowledge distillation / model transfer · WSI engineering · self-supervised learning · cell / tissue modeling · cross-resolution distillation · low-magnification histology · feature-map matching · unlabeled transfer set · compute reduction
Official artifacts: Open-access article · Author manuscript
Metadata checked 2026-08-23
2021ICMLPeer reviewed
Scaling Up Visual and Vision-Language Representation Learning With Noisy Text Supervision
Chao Jia et al.
L2
Image–text representation
image–text representation · ALIGN · dual encoder · noisy web pairs
Metadata checked 2026-07-27
2021ICLRPeer reviewed
Score-Based Generative Modeling through Stochastic Differential Equations
Yang Song et al.
L3
Generative / representation Generative vision Restoration / inverse problems
generative / representation learning · score model · stochastic differential equation · diffusion · generative vision
Metadata checked 2026-07-27
2021ICLRPeer reviewed
Sharpness-Aware Minimization for Efficiently Improving Generalization
Pierre Foret et al.
L2–L3
Neural networks
neural networks · SAM · sharpness · generalization
Metadata checked 2026-07-27
2021TACLPeer reviewed
SummEval: Re-evaluating Summarization Evaluation
Alexander R. Fabbri, Wojciech Kryściński, Bryan McCann, Caiming Xiong, Richard Socher, and Dragomir Radev
L2–L3
Generation / MT evaluation
generation / translation evaluation · SummEval · summarization evaluation · human judgments · metric meta-evaluation
Official artifacts: Official toolkit and annotations
Metadata checked 2026-08-10
2021ICCVPeer reviewed
Swin Transformer: Hierarchical Vision Transformer Using Shifted Windows
Ze Liu et al.
L2
Computer vision Visual representations Image / recognition
computer vision · vision transformer · hierarchy · visual representation learning · image recognition
Metadata checked 2026-07-27
2021ICCV WorkshopsPeer reviewed
SwinIR: Image Restoration Using Swin Transformer
Jingyun Liang et al.
L2
Generative vision Restoration / inverse problems
generative vision · SwinIR · image restoration · Swin Transformer
Metadata checked 2026-07-27
2021ICLRPeer reviewed
Tent: Fully Test-Time Adaptation by Entropy Minimization
Dequan Wang, Evan Shelhamer, Shaoteng Liu, Bruno Olshausen, and Trevor Darrell
L3
Published at ICLR 2021; OpenReview is the canonical conference record.
Transfer / adaptation / continual Evaluation / reproducibility Vision evaluation
evaluation / reproducibility · vision evaluation / robustness · TENT · test-time adaptation · entropy minimization · distribution shift · normalization
Official artifacts: Official implementation
Metadata checked 2026-08-30
2021ACL-IJCNLPPeer reviewed
Text2Event: Controllable Sequence-to-Structure Generation for End-to-end Event Extraction
Yaojie Lu, Hongyu Lin, Jin Xu, Xianpei Han, Jialong Tang, Annan Li, Le Sun, Meng Liao, and Shaoyi Chen
L2–L3
Structure / extraction
linguistic structure / information extraction · event extraction · sequence-to-structure generation · constrained decoding · curriculum learning
Official artifacts: Official implementation
Metadata checked 2026-08-10
2021ICMLPeer reviewed
Training Data-Efficient Image Transformers & Distillation through Attention
Hugo Touvron et al.
L2
Visual representations Image / recognition Knowledge distillation
visual representation learning · DeiT · vision transformer · distillation · image recognition · knowledge distillation / model transfer
Metadata checked 2026-07-27
2021EMNLPPeer reviewed
VideoCLIP: Contrastive Pre-training for Zero-shot Video-Text Understanding
Hu Xu et al.
L2
Generation / video / agents
multimodal generation / agents · video-text contrast · temporal positives · zero-shot
Metadata checked 2026-07-27
2021ICMLPeer reviewed
ViLT: Vision-and-Language Transformer Without Convolution or Region Supervision
Wonjae Kim et al.
L2
Image–text representation
image–text representation · patch tokens · single-stream fusion · region-free pretraining
Metadata checked 2026-07-28
2021CVPRPeer reviewed
VinVL: Revisiting Visual Representations in Vision-Language Models
Pengchuan Zhang et al.
L2
Captioning / VQA
captioning / VQA · region features · object attributes · captioning
Metadata checked 2026-07-27
2021ICCVPeer reviewed
Vision Transformers for Dense Prediction
René Ranftl, Alexey Bochkovskiy, and Vladlen Koltun
L2
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · DPT · depth estimation · dense prediction · 3D / spatial vision
Metadata checked 2026-07-27
2021ICMLPeer reviewed
WILDS: A Benchmark of in-the-Wild Distribution Shifts
Pang Wei Koh et al.
L2–L3
Evaluation / reproducibility NLP tasks / evaluation Vision evaluation
evaluation / reproducibility · WILDS · distribution shift · benchmark · NLP tasks / evaluation · vision evaluation / robustness · robustness / domain shift / privacy
Metadata checked 2026-07-27
2021ICMLPeer reviewed
Zero-Shot Text-to-Image Generation
Aditya Ramesh et al.
L2
Generation / video / agents
multimodal generation / agents · DALL-E · text-to-image · discrete visual tokens
Metadata checked 2026-07-27
2020ICMLPeer reviewed
A Simple Framework for Contrastive Learning of Visual Representations
Ting Chen et al.
L2
Computer vision Generative / representation Visual representations Loss / objectives
computer vision · self-supervised learning · SimCLR · contrastive learning · generative / representation learning · visual representation learning · loss / objective design
Metadata checked 2026-07-27
2020ACLPeer reviewed
BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehension
Mike Lewis et al.
L2
Transformers / pretraining Generation / MT evaluation
NLP transformers / pretraining · BART · denoising · encoder–decoder · generation / translation evaluation
Metadata checked 2026-07-27
2020ICLRPeer reviewed
BERTScore: Evaluating Text Generation with BERT
Tianyi Zhang et al.
L2
NLP tasks / evaluation Generation / MT evaluation
NLP tasks / evaluation · BERTScore · generation evaluation · semantic similarity · generation / translation evaluation
Metadata checked 2026-07-27
2020NeurIPSPeer reviewed
Big Bird: Transformers for Longer Sequences
Manzil Zaheer et al.
L2–L3
Transformers / pretraining
NLP transformers / pretraining · BigBird · sparse attention · long context
Metadata checked 2026-07-27
2020NeurIPSPeer reviewed
Bootstrap Your Own Latent
Jean-Bastien Grill et al.
Research
Generative / representation Visual representations Knowledge distillation
self-supervised learning · BYOL · self-distillation · generative / representation learning · visual representation learning · knowledge distillation / model transfer
Metadata checked 2026-07-27
2020Findings of EMNLPPeer reviewed
CodeBERT: A Pre-Trained Model for Programming and Natural Languages
Zhangyin Feng, Daya Guo, Duyu Tang, Nan Duan, Xiaocheng Feng, Ming Gong, Linjun Shou, Bing Qin, Ting Liu, Daxin Jiang, and Ming Zhou
L2–L3
Code / software engineering
code generation / software engineering · CodeBERT · code representation · MLM · replaced-token detection · code search
Official artifacts: Official implementation and checkpoints
Metadata checked 2026-08-20
2020SIGIRPeer reviewed
ColBERT: Efficient and Effective Passage Search via Contextualized Late Interaction over BERT
Omar Khattab and Matei Zaharia
L2
NLP tasks / evaluation
NLP tasks / evaluation · ColBERT · late interaction · passage retrieval
Official artifacts: Code
Metadata checked 2026-07-28
2020EMNLPPeer reviewed
COMET: A Neural Framework for MT Evaluation
Ricardo Rei, Craig Stewart, Ana C. Farinha, and Alon Lavie
L2–L3
Generation / MT evaluation
generation / translation evaluation · COMET · machine translation evaluation · learned metric
Official artifacts: Official code
Metadata checked 2026-08-08
2020NeurIPSPeer reviewed
Conservative Q-Learning for Offline Reinforcement Learning
Aviral Kumar et al.
L3
AI / reinforcement learning
AI / reinforcement learning · CQL · offline reinforcement learning · pessimism
Metadata checked 2026-07-27
2020ICLRPeer reviewed
Contrastive Representation Distillation
Yonglong Tian, Dilip Krishnan, and Phillip Isola
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · contrastive representation distillation · mutual information · teacher–student features
Official artifacts: Official code
Metadata checked 2026-08-10
2020NeurIPSPeer reviewed
Denoising Diffusion Probabilistic Models
Jonathan Ho, Ajay Jain, and Pieter Abbeel
L2
Generative / representation Generative vision Restoration / inverse problems
generative / representation learning · diffusion · denoising · score matching · generative vision · generative / virtual pathology
Metadata checked 2026-07-27
2020EMNLPPeer reviewed
Dense Passage Retrieval for Open-Domain Question Answering
Vladimir Karpukhin et al.
L2
NLP tasks / evaluation RAG / tools / agents
NLP tasks / evaluation · RAG / tools / agents · DPR · dense retrieval · question answering
Metadata checked 2026-07-27
2020CVPRPeer reviewed
Dreaming to Distill: Data-Free Knowledge Transfer via DeepInversion
Hongxu Yin, Pavlo Molchanov, Jose M. Alvarez, Zhizhong Li, Arun Mallya, Derek Hoiem, Niraj K. Jha, and Jan Kautz
L2–L3
Knowledge distillation Privacy / federated Visual representations
knowledge distillation / model transfer · privacy / federated learning · visual representation learning · data-free distillation · model inversion · batch-normalization statistics · adaptive deepinversion
Official artifacts: Official code
Metadata checked 2026-08-10
2020ICLRPeer reviewed
ELECTRA: Pre-training Text Encoders as Discriminators Rather Than Generators
Kevin Clark et al.
L2
Transformers / pretraining
NLP transformers / pretraining · ELECTRA · replaced-token detection · pretraining
Metadata checked 2026-07-27
2020CVPRPeer reviewed
End-to-End Learning of Visual Representations From Uncurated Instructional Videos
Antoine Miech et al.
L2
Generation / video / agents
multimodal generation / agents · MIL-NCE · video-text · temporal alignment
Metadata checked 2026-07-27
2020ECCVPeer reviewed
End-to-End Object Detection with Transformers
Nicolas Carion et al.
L2
Detection / segmentation
detection / segmentation · DETR · set prediction · object queries
Metadata checked 2026-07-27
2020JMLRPeer reviewed
Exploring the Limits of Transfer Learning with a Unified Text-to-Text Transformer
Colin Raffel et al.
L2
NLP Transformers / pretraining
NLP · transformer · text-to-text · NLP transformers / pretraining
Metadata checked 2026-07-27
2020MLSysPeer reviewed
Federated Optimization in Heterogeneous Networks
Tian Li, Anit Kumar Sahu, Manzil Zaheer, Maziar Sanjabi, Ameet Talwalkar, and Virginia Smith
L3
Published in the MLSys 2020 proceedings.
Privacy / federated
privacy / federated learning · FedProx · federated learning · heterogeneity · proximal objective · variable client work
Official artifacts: Official implementation
Metadata checked 2026-08-30
2020NeurIPSPeer reviewed
FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
Kihyuk Sohn et al.
L2
Generative / representation Transfer / adaptation / continual
generative / representation learning · FixMatch · semi-supervised learning · consistency regularization
Metadata checked 2026-07-27
2020arXivPreprint
GLU Variants Improve Transformer
Noam Shazeer
L2–L3
Neural networks
neural networks · SwiGLU · GEGLU · gated feed-forward network · Transformer
Metadata checked 2026-08-20
2020arXivTechnical report
Improved Baselines with Momentum Contrastive Learning
Xinlei Chen et al.
Research
Visual representations
self-supervised learning · MoCo v2 · contrastive learning · visual representation learning
Metadata checked 2026-07-27
2020AAAIPeer reviewed
Improved Knowledge Distillation via Teacher Assistant
Seyed Iman Mirzadeh, Mehrdad Farajtabar, Ang Li, Nir Levine, Akihiro Matsukawa, and Hassan Ghasemzadeh
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · teacher assistant · capacity gap · multistep distillation · model compression
Official artifacts: Official code
Metadata checked 2026-08-10
2020CVPRPeer reviewed
Iterative Answer Prediction With Pointer-Augmented Multimodal Transformers for TextVQA
Ronghang Hu et al.
L2
Captioning / VQA Document intelligence
captioning / VQA · TextVQA · pointer network · OCR · document intelligence
Metadata checked 2026-07-27
2020NeurIPSPeer reviewed
Language Models are Few-Shot Learners
Tom B. Brown et al.
L2
LLM pretraining
LLM pretraining / scaling · GPT-3 · in-context learning · scaling
Metadata checked 2026-07-27
2020KDDPeer reviewed
LayoutLM: Pre-training of Text and Layout for Document Image Understanding
Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, and Ming Zhou
L2–L3
Document intelligence
document intelligence · LayoutLM · 2D position embedding · OCR
Official artifacts: Official implementation and models
Metadata checked 2026-08-20
2020NeurIPSPeer reviewed
Learning to Summarize from Human Feedback
Nisan Stiennon et al.
L2
LLM post-training
LLM post-training / alignment · human feedback · reward model · summarization
Official artifacts: Code and data
Metadata checked 2026-07-28
2020arXivPreprint
Longformer: The Long-Document Transformer
Iz Beltagy, Matthew E. Peters, and Arman Cohan
L2
Transformers / pretraining
NLP transformers / pretraining · Longformer · sparse attention · long documents
Metadata checked 2026-07-27
2020EMNLPPeer reviewed
MAD-X: An Adapter-Based Framework for Multi-Task Cross-Lingual Transfer
Jonas Pfeiffer, Ivan Vulić, Iryna Gurevych, and Sebastian Ruder
L2–L3
Multilingual / low resource Transformers / pretraining Transfer / adaptation / continual
multilingual / low-resource NLP · NLP transformers / pretraining · MAD-X · language adapters · task adapters · cross-lingual transfer
Official artifacts: Official AdapterHub resource · Official adapters implementation
Metadata checked 2026-08-10
2020NaturePeer reviewed
Mastering Atari, Go, Chess and Shogi by Planning with a Learned Model
Julian Schrittwieser et al.
L2–L3
AI / reinforcement learning
AI / reinforcement learning · MuZero · model-based reinforcement learning · tree search
Metadata checked 2026-07-27
2020NeurIPSPeer reviewed
MiniLM: Deep Self-Attention Distillation for Task-Agnostic Compression of Pre-Trained Transformers
Wenhui Wang, Furu Wei, Li Dong, Hangbo Bao, Nan Yang, and Ming Zhou
L2–L3
Knowledge distillation Transformers / pretraining LLM systems
knowledge distillation / model transfer · NLP transformers / pretraining · LLM training / serving systems · self-attention distillation · value relations · transformer compression · task-agnostic pretraining
Official artifacts: Official implementation and checkpoints
Metadata checked 2026-08-10
2020CVPRPeer reviewed
Momentum Contrast for Unsupervised Visual Representation Learning
Kaiming He et al.
Research
Generative / representation Visual representations
self-supervised learning · MoCo · contrastive learning · generative / representation learning · visual representation learning
Metadata checked 2026-07-27
2020TACLPeer reviewed
Multilingual Denoising Pre-training for Neural Machine Translation
Yinhan Liu, Jiatao Gu, Naman Goyal, Xian Li, Sergey Edunov, Marjan Ghazvininejad, Mike Lewis, and Luke Zettlemoyer
L2–L3
Multilingual / low resource Generation / MT evaluation Transformers / pretraining
multilingual / low-resource NLP · generation / translation evaluation · NLP transformers / pretraining · mBART · multilingual denoising · sequence-to-sequence pretraining · machine translation
Official artifacts: Official fairseq implementation and checkpoints
Metadata checked 2026-08-10
2020ECCVPeer reviewed
NeRF: Representing Scenes as Neural Radiance Fields for View Synthesis
Ben Mildenhall et al.
L2–L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · NeRF · neural rendering · radiance field · 3D / spatial vision
Metadata checked 2026-07-27
2020ECCVPeer reviewed
OSCAR: Object-Semantics Aligned Pre-training for Vision-Language Tasks
Xiujun Li et al.
L1–L2
Image–text representation Captioning / VQA
image–text representation · captioning / VQA · object tags · fusion encoder · vision-language pretraining
Metadata checked 2026-07-28
2020ICMLPeer reviewed
PEGASUS: Pre-training with Extracted Gap-sentences for Abstractive Summarization
Jingqing Zhang, Yao Zhao, Mohammad Saleh, and Peter J. Liu
L2–L3
Generation / MT evaluation Transformers / pretraining
generation / translation evaluation · NLP transformers / pretraining · PEGASUS · gap-sentence generation · summarization
Official artifacts: Official code
Metadata checked 2026-08-08
2020ECCVPeer reviewed
RAFT: Recurrent All-Pairs Field Transforms for Optical Flow
Zachary Teed and Jia Deng
L2
Video / 3D / embodied
video / 3D / embodied vision · optical flow · all-pairs correlation · recurrent refinement
Official artifacts: Code
Metadata checked 2026-07-28
2020CVPR WorkshopsPeer reviewed
RandAugment: Practical Automated Data Augmentation with a Reduced Search Space
Ekin D. Cubuk et al.
L1–L2
Image / recognition
image recognition · RandAugment · augmentation · training recipe
Metadata checked 2026-07-27
2020ICMLPeer reviewed
REALM: Retrieval-Augmented Language Model Pre-Training
Kelvin Guu et al.
L2
RAG / tools / agents
RAG / tools / agents · REALM · retrieval-augmented pretraining
Metadata checked 2026-07-27
2020NeurIPSPeer reviewed
Residual Distillation: Towards Portable Deep Neural Networks without Shortcuts
Guilin Li, Junlei Zhang, Yunhe Wang, Chuanjian Liu, Matthias Tan, Yunfeng Lin, Wei Zhang, Jiashi Feng, and Tong Zhang
L3
Published in the NeurIPS 2020 main conference proceedings.
Knowledge distillation Neural networks Image / recognition Visual representations
knowledge distillation / model transfer · neural networks · image recognition · visual representation learning · residual distillation · plain CNN · shortcut removal · gradient blending · model deployment
Official artifacts: Official implementation
Metadata checked 2026-09-02
2020ECCVPeer reviewed
Resolution Switchable Networks for Runtime Efficient Image Recognition
Yikai Wang, Fuchun Sun, Duo Li, and Anbang Yao
L3
Peer-reviewed ECCV 2020 proceedings paper. One weight-shared CNN supports five trained input resolutions with private batch-normalization parameters; inference evaluates one chosen resolution and does not run the training-time ensemble.
Knowledge distillation Image / recognition Visual representations Vision evaluation
knowledge distillation / model transfer · image recognition · visual representation learning · vision evaluation / robustness · RS-Net · resolution switching · multi-resolution ensemble distillation · private batch normalization · runtime efficiency
Official artifacts: Official code and released models · Authors' arXiv manuscript
Metadata checked 2026-09-02
2020NeurIPSPeer reviewed
Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
Patrick Lewis et al.
L2
LLMs RAG / tools / agents
LLM · retrieval · generation · RAG / tools / agents
Metadata checked 2026-07-27
2020ICMLPeer reviewed
SCAFFOLD: Stochastic Controlled Averaging for Federated Learning
Sai Praneeth Karimireddy, Satyen Kale, Mehryar Mohri, Sashank Reddi, Sebastian Stich, and Ananda Theertha Suresh
L3
Published in the ICML 2020 proceedings, PMLR 119; no canonical first-party implementation was verified.
Privacy / federated
privacy / federated learning · SCAFFOLD · federated learning · client drift · control variates · heterogeneity
Metadata checked 2026-08-30
2020arXivPreprint
Scaling Laws for Neural Language Models
Jared Kaplan et al.
L2
LLMs LLM pretraining
LLM · scaling laws · LLM pretraining / scaling
Metadata checked 2026-07-27
2020TACLPeer reviewed
SpanBERT: Improving Pre-training by Representing and Predicting Spans
Mandar Joshi et al.
L2–L3
Structure / extraction Transformers / pretraining
linguistic structure / information extraction · NLP transformers / pretraining · SpanBERT · span masking · span boundary objective
Official artifacts: Official code and models
Metadata checked 2026-08-08
2020NeurIPSPeer reviewed
Supervised Contrastive Learning
Prannay Khosla et al.
L2
Generative / representation Visual representations Loss / objectives
generative / representation learning · visual representation learning · supervised contrastive learning · metric learning · representation learning · loss / objective design
Official artifacts: Code
Metadata checked 2026-07-28
2020Findings of EMNLPPeer reviewed
TinyBERT: Distilling BERT for Natural Language Understanding
Xiaoqi Jiao, Yichun Yin, Lifeng Shang, Xin Jiang, Xiao Chen, Linlin Li, Fang Wang, and Qun Liu
L2–L3
Knowledge distillation Transformers / pretraining
knowledge distillation / model transfer · NLP transformers / pretraining · TinyBERT · Transformer distillation · two-stage distillation
Official artifacts: Official implementation
Metadata checked 2026-08-10
2020ECCVPeer reviewed
UNITER: UNiversal Image-TExt Representation Learning
Yen-Chun Chen et al.
L2
Image–text representation
image–text representation · single-stream fusion · hard negatives · optimal transport
Metadata checked 2026-07-27
2020ACLPeer reviewed
Unsupervised Cross-lingual Representation Learning at Scale
Alexis Conneau et al.
L2
Transformers / pretraining Multilingual / low resource
NLP transformers / pretraining · XLM-R · multilingual pretraining · cross-lingual transfer · multilingual / low-resource NLP
Official artifacts: Models and code
Metadata checked 2026-07-28
2020NeurIPSPeer reviewed
Unsupervised Learning of Visual Features by Contrasting Cluster Assignments
Mathilde Caron et al.
Research
Visual representations
self-supervised learning · SwAV · prototypes · clustering · visual representation learning
Metadata checked 2026-07-27
2020ACLPeer reviewed
VisualBERT: A Simple and Performant Baseline for Vision and Language
Liunian Harold Li et al.
L2
Image–text representation
image–text representation · single-stream · region features · pretraining
Metadata checked 2026-07-27
2020SCPeer reviewed
ZeRO: Memory Optimizations Toward Training Trillion Parameter Models
Samyam Rajbhandari, Jeff Rasley, Olatunji Ruwase, and Yuxiong He
L2–L3
LLM systems LLM pretraining
LLM training / serving systems · LLM pretraining / scaling · ZeRO · state partitioning · distributed training
Official artifacts: DeepSpeed
Metadata checked 2026-08-08
2019CVPRPeer reviewed
A General and Adaptive Robust Loss Function
Jonathan T. Barron
L2–L3
Loss / objectives Vision evaluation
loss / objective design · vision evaluation / robustness · adaptive robust loss · M-estimation · learnable robustness · heavy-tailed distribution · outlier resistance
Official artifacts: Official PyTorch implementation
Metadata checked 2026-08-10
2019CVPRPeer reviewed
A Style-Based Generator Architecture for Generative Adversarial Networks
Tero Karras, Samuli Laine, and Timo Aila
L2–L3
Generative vision
generative vision · StyleGAN · GAN · style modulation
Metadata checked 2026-07-27
2019CVPRPeer reviewed
ArcFace: Additive Angular Margin Loss for Deep Face Recognition
Jiankang Deng, Jia Guo, Niannan Xue, and Stefanos Zafeiriou
L2–L3
Loss / objectives Visual representations
loss / objective design · visual representation learning · additive angular margin · face recognition · hyperspherical embedding · metric learning · normalized softmax
Official artifacts: Official implementation family
Metadata checked 2026-08-10
2019ICLRPeer reviewed
Benchmarking Neural Network Robustness to Common Corruptions and Perturbations
Dan Hendrycks and Thomas Dietterich
L2
Vision evaluation
vision evaluation / robustness · ImageNet-C · corruption robustness · distribution shift
Metadata checked 2026-07-27
2019NAACL-HLTPeer reviewed
BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding
Jacob Devlin et al.
L2
NLP Transformers / pretraining
NLP · transformer · pretraining · NLP transformers / pretraining
Metadata checked 2026-07-27
2019MIDLPeer reviewed
Boundary loss for highly unbalanced segmentation
Hoel Kervadec, Jihene Bouchtiba, Christian Desrosiers, Eric Granger, Jose Dolz, and Ismail Ben Ayed
L2–L3
Loss / objectives Detection / segmentation
loss / objective design · detection / segmentation · boundary loss · contour distance · level-set distance map · class imbalance · medical image segmentation
Official artifacts: Official code
Metadata checked 2026-08-10
2019CVPRPeer reviewed
Class-Balanced Loss Based on Effective Number of Samples
Yin Cui, Menglin Jia, Tsung-Yi Lin, Yang Song, and Serge Belongie
L2–L3
Loss / objectives Vision evaluation
loss / objective design · vision evaluation / robustness · class-balanced loss · effective number of samples · long-tailed recognition
Official artifacts: Official code
Metadata checked 2026-08-10
2019arXivTechnical report
Cross-Resolution Face Recognition via Prior-Aided Face Hallucination and Residual Knowledge Distillation
Hanyang Kong, Jian Zhao, Xiaoguang Tu, Junliang Xing, Shengmei Shen, and Jiashi Feng
L3
arXiv technical report submitted 26 May 2019; no peer-reviewed version or released implementation was located by 2026-09-02.
Knowledge distillation Image / recognition Generative vision Restoration / inverse problems Visual representations
knowledge distillation / model transfer · image recognition · generative vision · visual representation learning · RIM · cross-resolution face recognition · face hallucination · residual feature distillation · teacher assistant
Metadata checked 2026-09-02
2019ICCVPeer reviewed
CutMix: Regularization Strategy to Train Strong Classifiers with Localizable Features
Sangdoo Yun et al.
L2
Image / recognition
image recognition · CutMix · augmentation · regularization
Metadata checked 2026-07-27
2019ICCVPeer reviewed
Data-Free Learning of Student Networks
Hanting Chen, Yunhe Wang, Chang Xu, Zhaohui Yang, Chuanjian Liu, Boxin Shi, Chunjing Xu, Chao Xu, and Qi Tian
L2–L3
Peer-reviewed ICCV 2019 proceedings paper.
Knowledge distillation Privacy / federated
knowledge distillation / model transfer · privacy / federated learning · data-free distillation · generator · teacher decision boundary · student network · privacy boundary
Official artifacts: Author manuscript
Metadata checked 2026-08-23
2019ICLRPeer reviewed
Decoupled Weight Decay Regularization
Ilya Loshchilov and Frank Hutter
L2
Neural networks
neural networks · AdamW · weight decay · optimization
Metadata checked 2026-07-27
2019NeurIPSPeer reviewed
Deep Leakage from Gradients
Ligeng Zhu, Zhijian Liu, and Song Han
L2–L3
Privacy / federated
privacy / federated learning · gradient inversion · federated learning · privacy attack · shared gradients · data reconstruction · robustness / domain shift / privacy
Official artifacts: Final proceedings PDF · Authors' implementation
Metadata checked 2026-08-20
2019NeurIPS EMC² WorkshopTechnical report
DistilBERT, a Distilled Version of BERT: Smaller, Faster, Cheaper and Lighter
Victor Sanh, Lysandre Debut, Julien Chaumond, and Thomas Wolf
L2–L3
Workshop paper and arXiv preprint; not a main-track NeurIPS proceedings paper.
Knowledge distillation Transformers / pretraining LLM systems
knowledge distillation / model transfer · NLP transformers / pretraining · LLM training / serving systems · DistilBERT · pretraining distillation · masked language modeling · cosine loss · Transformer compression
Official artifacts: Official distillation code
Metadata checked 2026-08-23
2019ICMLPeer reviewed
Do ImageNet Classifiers Generalize to ImageNet?
Benjamin Recht et al.
L2
Vision evaluation
vision evaluation / robustness · ImageNetV2 · test-set replication · distribution shift
Metadata checked 2026-07-27
2019ICMLPeer reviewed
EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks
Mingxing Tan and Quoc V. Le
L2
Image / recognition
image recognition · EfficientNet · compound scaling · efficiency
Metadata checked 2026-07-27
2019ICCVPeer reviewed
FCOS: Fully Convolutional One-Stage Object Detection
Zhi Tian, Chunhua Shen, Hao Chen, and Tong He
L2–L3
Detection / segmentation
detection / segmentation · anchor-free detection · one-stage detection · center-ness · feature pyramid network
Official artifacts: Official implementation
Metadata checked 2026-08-10
2019CVPRPeer reviewed
Generalized Intersection Over Union: A Metric and A Loss for Bounding Box Regression
Hamid Rezatofighi, Nathan Tsoi, JunYoung Gwak, Amir Sadeghian, Ian Reid, and Silvio Savarese
L2–L3
Loss / objectives Detection / segmentation
loss / objective design · detection / segmentation · generalized IoU · bounding box regression · localization metric · object detection · non-overlapping boxes
Official artifacts: Official project and implementations
Metadata checked 2026-08-10
2019NeurIPSPeer reviewed
Generative Modeling by Estimating Gradients of the Data Distribution
Yang Song and Stefano Ermon
L3
Generative / representation
generative / representation learning · noise-conditioned score network · score matching · Langevin dynamics
Metadata checked 2026-07-27
2019ICLRPeer reviewed
GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding
Alex Wang et al.
L1–L2
NLP tasks / evaluation
NLP tasks / evaluation · GLUE · natural language understanding · benchmark
Metadata checked 2026-07-27
2019CVPRPeer reviewed
GQA: A New Dataset for Real-World Visual Reasoning and Compositional Question Answering
Drew A. Hudson and Christopher D. Manning
L2
Captioning / VQA V&L evaluation / safety
captioning / VQA · vision–language evaluation / safety · compositional reasoning · scene graphs · functional programs
Metadata checked 2026-07-27
2019ICLRPeer reviewed
How Powerful are Graph Neural Networks?
Keyulu Xu, Weihua Hu, Jure Leskovec, and Stefanie Jegelka
L2–L3
Graph / geometric ML Neural networks
graph / geometric learning · neural networks · GIN · Weisfeiler–Lehman test · GNN expressivity
Official artifacts: Official code
Metadata checked 2026-08-08
2019ICCVPeer reviewed
HowTo100M: Learning a Text-Video Embedding by Watching Hundred Million Narrated Video Clips
Antoine Miech et al.
L2
Generation / video / agents
multimodal generation / agents · video-text · weak supervision · large-scale data
Metadata checked 2026-07-27
2019ICLRPeer reviewed
ImageNet-Trained CNNs Are Biased towards Texture; Increasing Shape Bias Improves Accuracy and Robustness
Robert Geirhos et al.
L2
Vision evaluation
vision evaluation / robustness · texture bias · shape bias · robustness
Metadata checked 2026-07-27
2019NeurIPSPeer reviewed
Learning Imbalanced Datasets with Label-Distribution-Aware Margin Loss
Kaidi Cao, Colin Wei, Adrien Gaidon, Nikos Arechiga, and Tengyu Ma
L2–L3
Loss / objectives Image / recognition
loss / objective design · image recognition · long-tailed recognition · class-dependent margin · deferred re-weighting · label imbalance · minority classes
Official artifacts: Official code
Metadata checked 2026-08-10
2019EMNLP-IJCNLPPeer reviewed
LXMERT: Learning Cross-Modality Encoder Representations from Transformers
Hao Tan and Mohit Bansal
L2
Image–text representation
image–text representation · cross-modal encoder · region features · pretraining
Metadata checked 2026-07-27
2019FAT* / FAccTPeer reviewed
Model Cards for Model Reporting
Margaret Mitchell et al.
L1–L2
Evaluation / reproducibility LLM evaluation / safety Vision evaluation
evaluation / reproducibility · model cards · documentation · responsible AI · LLM evaluation / safety · vision evaluation / robustness
Metadata checked 2026-07-27
2019ICCVPeer reviewed
nocaps: novel object captioning at scale
Harsh Agrawal et al.
L2
Captioning / VQA
captioning / VQA · captioning · novel objects · benchmark
Metadata checked 2026-07-27
2019NeurIPSPeer reviewed
ObjectNet: A Large-Scale Bias-Controlled Dataset for Pushing the Limits of Object Recognition Models
Andrei Barbu, David Mayo, Julian Alverio, William Luo, Christopher Wang, Dan Gutfreund, Josh Tenenbaum, and Boris Katz
L2–L3
Vision evaluation
vision evaluation / robustness · ObjectNet · bias-controlled dataset · distribution shift · object recognition evaluation
Official artifacts: Official dataset website · Official evaluation repository · Semantic Scholar citation record
Metadata checked 2026-08-20
2019ICDARPeer reviewed
OCR-VQA: Visual Question Answering by Reading Text in Images
Anand Mishra et al.
L2
Captioning / VQA Document intelligence
captioning / VQA · OCR · VQA · dataset · document intelligence
Metadata checked 2026-07-27
2019CVPRPeer reviewed
OK-VQA: A Visual Question Answering Benchmark Requiring External Knowledge
Kenneth Marino et al.
L2
Captioning / VQA
captioning / VQA · knowledge VQA · external knowledge · open-ended answers
Metadata checked 2026-07-27
2019CVPRPeer reviewed
Panoptic Feature Pyramid Networks
Alexander Kirillov et al.
L2
Detection / segmentation
detection / segmentation · panoptic segmentation · FPN · things and stuff
Metadata checked 2026-07-27
2019ICMLPeer reviewed
Parameter-Efficient Transfer Learning for NLP
Neil Houlsby et al.
L2
Transformers / pretraining LLM post-training Transfer / adaptation / continual
NLP transformers / pretraining · LLM post-training / alignment · adapters · parameter-efficient tuning
Metadata checked 2026-07-27
2019PNASPeer reviewed
Reconciling Modern Machine-Learning Practice and the Classical Bias–Variance Trade-Off
Mikhail Belkin et al.
L2–L3
Neural networks
neural networks · double descent · bias variance · overparameterization
Metadata checked 2026-07-27
2019CVPRPeer reviewed
Relational Knowledge Distillation
Wonpyo Park, Dongju Kim, Yan Lu, and Minsu Cho
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · relational knowledge distillation · distance-wise loss · angle-wise loss
Official artifacts: Official code
Metadata checked 2026-08-10
2019arXivPreprint
RoBERTa: A Robustly Optimized BERT Pretraining Approach
Yinhan Liu et al.
L2
Transformers / pretraining
NLP transformers / pretraining · RoBERTa · masked language modeling · training recipe
Metadata checked 2026-07-27
2019NeurIPSPeer reviewed
Root Mean Square Layer Normalization
Biao Zhang and Rico Sennrich
L2
Transformers / pretraining LLM pretraining
NLP transformers / pretraining · LLM pretraining / scaling · RMSNorm · normalization · optimization
Metadata checked 2026-07-28
2019ICCVPeer reviewed
SlowFast Networks for Video Recognition
Christoph Feichtenhofer et al.
L2
Video / 3D / embodied
video / 3D / embodied vision · SlowFast · video recognition · temporal sampling
Metadata checked 2026-07-27
2019ICCVPeer reviewed
Symmetric Cross Entropy for Robust Learning with Noisy Labels
Yisen Wang, Xingjun Ma, Zaiyi Chen, Yuan Luo, Jinfeng Yi, and James Bailey
L2–L3
Loss / objectives Vision evaluation
loss / objective design · vision evaluation / robustness · symmetric cross entropy · reverse cross entropy · noisy labels · robust classification · hard-class learning
Official artifacts: Official code
Metadata checked 2026-08-10
2019CVPRPeer reviewed
Towards VQA Models That Can Read
Amanpreet Singh et al.
L2
Captioning / VQA Document intelligence
captioning / VQA · TextVQA · OCR · scene text · document intelligence
Metadata checked 2026-07-27
2019ACLPeer reviewed
Transformer-XL: Attentive Language Models beyond a Fixed-Length Context
Zihang Dai et al.
L2
Transformers / pretraining RAG / tools / agents
NLP transformers / pretraining · RAG / tools / agents · Transformer-XL · recurrence · long context
Metadata checked 2026-07-27
2019ICCVPeer reviewed
VideoBERT: A Joint Model for Video and Language Representation Learning
Chen Sun et al.
L2
Generation / video / agents
multimodal generation / agents · video-language pretraining · discrete visual tokens
Metadata checked 2026-07-27
2019NeurIPSPeer reviewed
ViLBERT: Pretraining Task-Agnostic Visiolinguistic Representations for Vision-and-Language Tasks
Jiasen Lu et al.
L1–L2
Image–text representation
image–text representation · two-stream fusion · co-attention · region features
Metadata checked 2026-07-27
2019ICMLPeer reviewed
Zero-Shot Knowledge Distillation in Deep Networks
Gaurav Kumar Nayak, Konda Reddy Mopuri, Vaisakh Shaj, R. Venkatesh Babu, and Anirban Chakraborty
L2–L3
Knowledge distillation Privacy / federated Visual representations
knowledge distillation / model transfer · privacy / federated learning · visual representation learning · zero-shot knowledge distillation · data-free distillation · data impressions · Dirichlet sampling · model inversion
Metadata checked 2026-08-10
2018Journal of Machine Learning ResearchPeer reviewed
Automatic Differentiation in Machine Learning: a Survey
Atilim Gunes Baydin, Barak A. Pearlmutter, Alexey Andreyevich Radul, and Jeffrey Mark Siskind
L1–L2
Math / statistics Neural networks
mathematics / statistics · neural networks · automatic differentiation · backpropagation · computation graphs
Metadata checked 2026-07-28
2018ICMLPeer reviewed
Born Again Neural Networks
Tommaso Furlanello, Zachary C. Lipton, Michael Tschannen, Laurent Itti, and Anima Anandkumar
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · born-again networks · self-distillation · teacher generations
Metadata checked 2026-08-10
2018CVPRPeer reviewed
Bottom-Up and Top-Down Attention for Image Captioning and Visual Question Answering
Peter Anderson et al.
L2
Captioning / VQA
captioning / VQA · region features · captioning · VQA
Metadata checked 2026-07-27
2018NeurIPSPeer reviewed
DAGs with NO TEARS: Continuous Optimization for Structure Learning
Xun Zheng, Bryon Aragam, Pradeep K. Ravikumar, and Eric P. Xing
L2–L3
Causal inference
causal inference · NOTEARS · causal discovery · DAG structure learning · continuous optimization
Official artifacts: Official code
Metadata checked 2026-08-08
2018ECCVPeer reviewed
Deep Clustering for Unsupervised Learning of Visual Features
Mathilde Caron et al.
Research
Visual representations
self-supervised learning · clustering · pseudo-labels · visual representation learning
Metadata checked 2026-07-27
2018NAACL-HLTPeer reviewed
Deep Contextualized Word Representations
Matthew E. Peters et al.
L1–L2
Text / sequence models
text / sequence modeling · ELMo · contextual embeddings · bidirectional language model
Metadata checked 2026-07-27
2018CVPRPeer reviewed
Deep Image Prior
Dmitry Ulyanov, Andrea Vedaldi, and Victor Lempitsky
L2–L3
Restoration / inverse problems
Deep Image Prior · untrained convolutional network · image restoration · architectural prior
Official artifacts: Official project and supplementary material · Official implementation · arXiv record
Metadata checked 2026-08-20
2018CVPRPeer reviewed
Deep Mutual Learning
Ying Zhang, Tao Xiang, Timothy M. Hospedales, and Huchuan Lu
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · deep mutual learning · online distillation · peer networks
Official artifacts: Official code
Metadata checked 2026-08-10
2018AAAIPeer reviewed
DeepHit: A Deep Learning Approach to Survival Analysis with Competing Risks
Changhee Lee, William R. Zame, Jinsung Yoon, and Mihaela van der Schaar
L2–L3
Loss / objectives
loss / objective design · survival analysis · competing risks · discrete event-time distribution · ranking loss · non-proportional hazards
Official artifacts: Official code
Metadata checked 2026-08-10
2018The Econometrics JournalPeer reviewed
Double/debiased machine learning for treatment and structural parameters
Victor Chernozhukov et al.
L2–L3
Causal inference Evaluation / reproducibility
causal inference · evaluation / reproducibility · orthogonal scores · cross-fitting · treatment effects
Official artifacts: DoubleML
Metadata checked 2026-08-08
2018ECCVPeer reviewed
Encoder-Decoder with Atrous Separable Convolution for Semantic Image Segmentation
Liang-Chieh Chen et al.
L2
Detection / segmentation
detection / segmentation · DeepLabv3+ · semantic segmentation · atrous convolution
Metadata checked 2026-07-27
2018Journal of the American Statistical AssociationPeer reviewed
Estimation and Inference of Heterogeneous Treatment Effects using Random Forests
Stefan Wager and Susan Athey
L2–L3
Causal inference
causal inference · causal forests · heterogeneous treatment effects · honest estimation
Official artifacts: grf implementation
Metadata checked 2026-08-08
2018Conference on Fairness, Accountability and TransparencyPeer reviewed
Gender Shades: Intersectional Accuracy Disparities in Commercial Gender Classification
Joy Buolamwini and Timnit Gebru
L2
Vision evaluation Evaluation / reproducibility
vision evaluation / robustness · intersectional evaluation · fairness · facial analysis · evaluation / reproducibility
Metadata checked 2026-07-27
2018NeurIPSPeer reviewed
Generalized Cross Entropy Loss for Training Deep Neural Networks with Noisy Labels
Zhilu Zhang and Mert R. Sabuncu
L2–L3
Loss / objectives Vision evaluation
loss / objective design · vision evaluation / robustness · generalized cross entropy · noisy labels · robust loss
Metadata checked 2026-08-10
2018NeurIPSPeer reviewed
Glow: Generative Flow with Invertible 1×1 Convolutions
Diederik P. Kingma and Prafulla Dhariwal
L2–L3
Generative / representation
generative / representation learning · Glow · normalizing flow · invertible convolution
Metadata checked 2026-07-27
2018ICLRPeer reviewed
Graph Attention Networks
Petar Veličković et al.
L2
Neural networks Graph / geometric ML
neural networks · graph attention network · message passing · attention · graph / geometric learning
Metadata checked 2026-07-27
2018ECCVPeer reviewed
Group Normalization
Yuxin Wu and Kaiming He
L2–L3
Neural networks
neural networks · GroupNorm · normalization · small-batch training · computer vision
Official artifacts: Official Detectron implementation
Metadata checked 2026-08-20
2018OpenAI technical reportTechnical report
Improving Language Understanding by Generative Pre-Training
Alec Radford et al.
L2
Transformers / pretraining
NLP transformers / pretraining · GPT · generative pretraining · transfer learning
Metadata checked 2026-07-27
2018ICMLPeer reviewed
Interpretability Beyond Feature Attribution: Quantitative Testing with Concept Activation Vectors (TCAV)
Been Kim, Martin Wattenberg, Justin Gilmer, Carrie Cai, James Wexler, Fernanda Viégas, and Rory Sayres
L3
Published in the ICML 2018 proceedings, PMLR 80.
Interpretability
interpretability / attribution · TCAV · concept activation vector · human concepts · directional derivative · interpretability
Official artifacts: Official implementation
Metadata checked 2026-08-30
2018ICLRPeer reviewed
Mixed Precision Training
Paulius Micikevicius et al.
L2
Neural networks LLM systems
neural networks · mixed precision · loss scaling · training systems · LLM training / serving systems
Metadata checked 2026-07-27
2018ICLRPeer reviewed
mixup: Beyond Empirical Risk Minimization
Hongyi Zhang et al.
L2
Image / recognition
image recognition · mixup · augmentation · regularization
Metadata checked 2026-07-27
2018NeurIPSPeer reviewed
Neural Ordinary Differential Equations
Ricky T. Q. Chen et al.
L3
Generative / representation Generative vision
generative / representation learning · neural ODE · continuous depth · continuous normalizing flow · generative vision
Metadata checked 2026-07-27
2018NeurIPSPeer reviewed
Neural Tangent Kernel: Convergence and Generalization in Neural Networks
Arthur Jacot, Franck Gabriel, and Clément Hongler
L3
Neural networks
neural networks · neural tangent kernel · optimization theory · infinite width
Metadata checked 2026-07-27
2018EMNLPPeer reviewed
Object Hallucination in Image Captioning
Anna Rohrbach et al.
L2
V&L evaluation / safety
vision–language evaluation / safety · CHAIR · caption hallucination · object factuality
Metadata checked 2026-07-27
2018arXivPreprint
Representation Learning with Contrastive Predictive Coding
Aaron van den Oord, Yazhe Li, and Oriol Vinyals
Research
Visual representations Loss / objectives
self-supervised learning · CPC · InfoNCE · visual representation learning · loss / objective design
Metadata checked 2026-07-27
2018NeurIPSPeer reviewed
Sanity Checks for Saliency Maps
Julius Adebayo, Justin Gilmer, Michael Muelly, Ian Goodfellow, Moritz Hardt, and Been Kim
L2–L3
Interpretability
interpretability / attribution · saliency maps · model randomization · data randomization · faithfulness · interpretability evaluation
Official artifacts: Official code and errata · Official supplement
Metadata checked 2026-08-20
2018EMNLP System DemonstrationsPeer reviewed
SentencePiece: A Simple and Language Independent Subword Tokenizer and Detokenizer for Neural Text Processing
Taku Kudo and John Richardson
L1
Text / sequence models
text / sequence modeling · SentencePiece · tokenization · multilingual
Metadata checked 2026-07-27
2018ICMLPeer reviewed
Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actor
Tuomas Haarnoja, Aurick Zhou, Pieter Abbeel, and Sergey Levine
L2
AI / reinforcement learning
AI / reinforcement learning · soft actor critic · maximum entropy RL · off-policy learning
Metadata checked 2026-07-28
2018CVPRPeer reviewed
The Lovász-Softmax Loss: A Tractable Surrogate for the Optimization of the Intersection-Over-Union Measure in Neural Networks
Maxim Berman, Amal Rannen Triki, and Matthew B. Blaschko
L2–L3
Loss / objectives Detection / segmentation
loss / objective design · detection / segmentation · Lovász-Softmax · intersection over union · semantic segmentation
Official artifacts: Official code
Metadata checked 2026-08-10
2018ICLRPeer reviewed
Towards Deep Learning Models Resistant to Adversarial Attacks
Aleksander Madry, Aleksandar Makelov, Ludwig Schmidt, Dimitris Tsipras, and Adrian Vladu
L2
Vision evaluation
vision evaluation / robustness · adversarial robustness · projected gradient descent · robust optimization
Metadata checked 2026-07-28
2018Journal of Open Source SoftwarePeer reviewed
UMAP: Uniform Manifold Approximation and Projection
Leland McInnes, John Healy, and James Melville
L2
Classical ML
classical machine learning · UMAP · visualization · manifold learning
Metadata checked 2026-07-27
2018ACLPeer reviewed
Universal Language Model Fine-tuning for Text Classification
Jeremy Howard and Sebastian Ruder
L1–L2
Transformers / pretraining Transfer / adaptation / continual
NLP transformers / pretraining · ULMFiT · fine-tuning · transfer learning
Metadata checked 2026-07-27
2018CVPRPeer reviewed
Unsupervised Feature Learning via Non-Parametric Instance Discrimination
Zhirong Wu et al.
Research
Visual representations
self-supervised learning · instance discrimination · contrastive learning · visual representation learning
Metadata checked 2026-07-27
2018ICLRPeer reviewed
Unsupervised Representation Learning by Predicting Image Rotations
Spyros Gidaris, Praveer Singh, and Nikos Komodakis
Research
Visual representations
self-supervised learning · pretext task · rotation · visual representation learning
Metadata checked 2026-07-27
2018CVPRPeer reviewed
VizWiz Grand Challenge: Answering Visual Questions from Blind People
Danna Gurari et al.
L2
Captioning / VQA V&L evaluation / safety
captioning / VQA · vision–language evaluation / safety · visual question answering · accessibility · unanswerable questions
Metadata checked 2026-07-28
2018BMVCPeer reviewed
VSE++: Improving Visual-Semantic Embeddings with Hard Negatives
Fartash Faghri, David J. Fleet, Jamie Ryan Kiros, and Sanja Fidler
L2–L3
Image–text representation Multimodal retrieval
image–text representation · multimodal embedding / retrieval · VSE++ · visual-semantic embedding · hard-negative mining · cross-modal retrieval
Official artifacts: Official implementation
Metadata checked 2026-08-20
2018EMNLPPeer reviewed
XNLI: Evaluating Cross-lingual Sentence Representations
Alexis Conneau et al.
L2–L3
Multilingual / low resource
multilingual / low-resource NLP · XNLI · cross-lingual transfer · natural language inference
Official artifacts: Official dataset
Metadata checked 2026-08-08
2017ICLRPeer reviewed
A Baseline for Detecting Misclassified and Out-of-Distribution Examples in Neural Networks
Dan Hendrycks and Kevin Gimpel
L2
Evaluation / reproducibility Vision evaluation
evaluation / reproducibility · vision evaluation / robustness · out-of-distribution detection · confidence · selective prediction
Metadata checked 2026-07-28
2017NeurIPSPeer reviewed
A Unified Approach to Interpreting Model Predictions
Scott M. Lundberg and Su-In Lee
L2–L3
Interpretability Evaluation / reproducibility LLM interpretability
interpretability / attribution · evaluation / reproducibility · SHAP · Shapley values · local explanations · LLM interpretability / editing / unlearning
Official artifacts: Official implementation
Metadata checked 2026-08-08
2017NeurIPSPeer reviewed
Attention Is All You Need
Ashish Vaswani et al.
L2
NLP Neural networks Transformers / pretraining LLM pretraining
NLP · transformer · attention · neural networks · NLP transformers / pretraining · LLM pretraining / scaling
Metadata checked 2026-07-27
2017ICMLPeer reviewed
Axiomatic Attribution for Deep Networks
Mukund Sundararajan, Ankur Taly, and Qiqi Yan
L2–L3
Interpretability Evaluation / reproducibility LLM interpretability
interpretability / attribution · evaluation / reproducibility · integrated gradients · sensitivity · implementation invariance · LLM interpretability / editing / unlearning
Metadata checked 2026-08-08
2017IEEE Transactions on Image ProcessingPeer reviewed
Beyond a Gaussian Denoiser: Residual Learning of Deep CNN for Image Denoising
Kai Zhang et al.
L2
Generative vision Restoration / inverse problems
generative vision · DnCNN · denoising · residual prediction
Metadata checked 2026-07-27
2017CVPRPeer reviewed
CLEVR: A Diagnostic Dataset for Compositional Language and Elementary Visual Reasoning
Justin Johnson et al.
L1–L2
Captioning / VQA V&L evaluation / safety
captioning / VQA · vision–language evaluation / safety · compositional reasoning · functional programs · diagnostic data
Metadata checked 2026-07-28
2017AISTATSPeer reviewed
Communication-Efficient Learning of Deep Networks from Decentralized Data
H. Brendan McMahan, Eider Moore, Daniel Ramage, Seth Hampson, and Blaise Agüera y Arcas
L2–L3
Privacy / federated Neural networks
privacy / federated learning · neural networks · FedAvg · federated learning · non-IID clients · robustness / domain shift / privacy
Metadata checked 2026-08-08
2017CVPRPeer reviewed
Densely Connected Convolutional Networks
Gao Huang et al.
L2
Image / recognition
image recognition · DenseNet · CNN · feature reuse
Metadata checked 2026-07-27
2017ICLRPeer reviewed
Density Estimation using Real NVP
Laurent Dinh, Jascha Sohl-Dickstein, and Samy Bengio
L2–L3
Generative / representation Generative vision
generative / representation learning · normalizing flow · exact likelihood · invertible network · generative vision
Metadata checked 2026-07-27
2017EMNLPPeer reviewed
End-to-end Neural Coreference Resolution
Kenton Lee, Luheng He, Mike Lewis, and Luke Zettlemoyer
L2–L3
Structure / extraction
linguistic structure / information extraction · coreference resolution · span ranking · end-to-end information extraction · structured prediction
Official artifacts: Official implementation
Metadata checked 2026-08-10
2017CVPR WorkshopsPeer reviewed
Enhanced Deep Residual Networks for Single Image Super-Resolution
Bee Lim et al.
L2
Generative vision Restoration / inverse problems
generative vision · EDSR · super-resolution · residual network
Metadata checked 2026-07-27
2017TACLPeer reviewed
Enriching Word Vectors with Subword Information
Piotr Bojanowski et al.
L1
Text / sequence models
text / sequence modeling · fastText · subwords · morphology
Metadata checked 2026-07-27
2017CVPRPeer reviewed
Feature Pyramid Networks for Object Detection
Tsung-Yi Lin et al.
L2
Detection / segmentation
detection / segmentation · FPN · object detection · multiscale features
Metadata checked 2026-07-27
2017ICCVPeer reviewed
Focal Loss for Dense Object Detection
Tsung-Yi Lin et al.
L2
Detection / segmentation Loss / objectives
detection / segmentation · RetinaNet · focal loss · one-stage detection · loss / objective design
Metadata checked 2026-07-27
2017NeurIPSPeer reviewed
GANs Trained by a Two Time-Scale Update Rule Converge to a Local Nash Equilibrium
Martin Heusel, Hubert Ramsauer, Thomas Unterthiner, Bernhard Nessler, and Sepp Hochreiter
L2
Generative / representation Evaluation / reproducibility Generative vision Vision evaluation
generative / representation learning · evaluation / reproducibility · generative vision · vision evaluation / robustness · Fréchet Inception Distance · GAN evaluation · two time-scale updates · generative / virtual pathology
Metadata checked 2026-07-28
2017DLMIAPeer reviewed
Generalised Dice overlap as a deep learning loss function for highly unbalanced segmentations
Carole H. Sudre, Wenqi Li, Tom Vercauteren, Sébastien Ourselin, and M. Jorge Cardoso
L2–L3
Loss / objectives Detection / segmentation
loss / objective design · detection / segmentation · generalised Dice loss · class imbalance · medical image segmentation · overlap objective · inverse-volume weighting
Official artifacts: Open author manuscript
Metadata checked 2026-08-10
2017ICCVPeer reviewed
Grad-CAM: Visual Explanations From Deep Networks via Gradient-Based Localization
Ramprasaath R. Selvaraju, Michael Cogswell, Abhishek Das, Ramakrishna Vedantam, Devi Parikh, and Dhruv Batra
L2–L3
Interpretability
interpretability / attribution · Grad-CAM · visual explanations · gradient localization
Official artifacts: Official code
Metadata checked 2026-08-08
2017CVPRPeer reviewed
Image-to-Image Translation with Conditional Adversarial Networks
Phillip Isola, Jun-Yan Zhu, Tinghui Zhou, and Alexei A. Efros
L3
Generative vision Restoration / inverse problems
generative vision · generative / virtual pathology · pix2pix · conditional GAN · paired image translation · PatchGAN
Official artifacts: Official implementation
Metadata checked 2026-08-10
2017NeurIPSPeer reviewed
Inductive Representation Learning on Large Graphs
William L. Hamilton, Rex Ying, and Jure Leskovec
L2
Neural networks Graph / geometric ML
neural networks · GraphSAGE · graph neural network · inductive learning · graph / geometric learning
Metadata checked 2026-07-27
2017ICMLPeer reviewed
Language Modeling with Gated Convolutional Networks
Yann N. Dauphin, Angela Fan, Michael Auli, and David Grangier
L2–L3
Neural networks
neural networks · GLU · gated linear unit · gated convolution · language modeling
Metadata checked 2026-08-20
2017CVPRPeer reviewed
Making the V in VQA Matter: Elevating the Role of Image Understanding in Visual Question Answering
Yash Goyal et al.
L1–L2
Captioning / VQA V&L evaluation / safety
captioning / VQA · vision–language evaluation / safety · VQA v2 · complementary images · language shortcuts
Metadata checked 2026-07-27
2017ICCVPeer reviewed
Mask R-CNN
Kaiming He, Georgia Gkioxari, Piotr Dollár, and Ross Girshick
L2
Computer vision Detection / segmentation
computer vision · instance segmentation · object detection · detection / segmentation
Metadata checked 2026-07-27
2017NaturePeer reviewed
Mastering the Game of Go without Human Knowledge
David Silver et al.
L2–L3
AI / reinforcement learning
AI / reinforcement learning · AlphaGo Zero · Monte Carlo tree search · self-play
Metadata checked 2026-07-27
2017IEEE Symposium on Security and PrivacyPeer reviewed
Membership Inference Attacks Against Machine Learning Models
Reza Shokri, Marco Stronati, Congzheng Song, and Vitaly Shmatikov
L2–L3
Privacy / federated
privacy / federated learning · membership inference · privacy attacks · shadow models
Metadata checked 2026-08-08
2017arXivTechnical report
MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications
Andrew G. Howard, Menglong Zhu, Bo Chen, Dmitry Kalenichenko, Weijun Wang, Tobias Weyand, Marco Andreetto, and Hartwig Adam
L3
Canonical 2017 arXiv technical report; no peer-reviewed proceedings version was located.
Image / recognition Neural networks
image recognition · neural networks · MobileNet V1 · depthwise separable convolution · width multiplier · resolution multiplier · mobile vision
Official artifacts: Official TensorFlow implementation and checkpoints
Metadata checked 2026-08-30
2017ICMLPeer reviewed
Model-Agnostic Meta-Learning for Fast Adaptation of Deep Networks
Chelsea Finn, Pieter Abbeel, and Sergey Levine
L2–L3
Generative / representation Transfer / adaptation / continual
generative / representation learning · MAML · meta-learning · few-shot adaptation
Metadata checked 2026-07-27
2017NeurIPSPeer reviewed
Neural Discrete Representation Learning
Aaron van den Oord, Oriol Vinyals, and Koray Kavukcuoglu
L2
Generative vision Generation / video / agents
generative vision · VQ-VAE · discrete latent · codebook · multimodal generation / agents
Metadata checked 2026-07-27
2017ICMLPeer reviewed
Neural Message Passing for Quantum Chemistry
Justin Gilmer, Samuel S. Schoenholz, Patrick F. Riley, Oriol Vinyals, and George E. Dahl
L2
Neural networks Graph / geometric ML
neural networks · message passing · graph neural networks · invariance · graph / geometric learning
Metadata checked 2026-07-28
2017ICMLPeer reviewed
On Calibration of Modern Neural Networks
Chuan Guo et al.
L2
Evaluation / reproducibility Vision evaluation Loss / objectives
evaluation / reproducibility · calibration · expected calibration error · temperature scaling · vision evaluation / robustness · loss / objective design
Metadata checked 2026-07-27
2017Proceedings of the National Academy of SciencesPeer reviewed
Overcoming Catastrophic Forgetting in Neural Networks
James Kirkpatrick et al.
L2
Generative / representation Transfer / adaptation / continual
generative / representation learning · elastic weight consolidation · continual learning · forgetting
Metadata checked 2026-07-28
2017ICLRPeer reviewed
Paying More Attention to Attention: Improving the Performance of Convolutional Neural Networks via Attention Transfer
Sergey Zagoruyko and Nikos Komodakis
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · attention transfer · activation maps · feature distillation
Official artifacts: Official code
Metadata checked 2026-08-10
2017CVPRPeer reviewed
PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation
Charles R. Qi et al.
L2
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · PointNet · point cloud · 3D representation · 3D / spatial vision
Metadata checked 2026-07-27
2017NeurIPSPeer reviewed
PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space
Charles Ruizhongtai Qi, Li Yi, Hao Su, and Leonidas Guibas
L2–L3
3D / spatial vision Video / 3D / embodied
3D / spatial vision · video / 3D / embodied vision · PointNet++ · point clouds · set abstraction · 3D representation learning
Official artifacts: Official implementation
Metadata checked 2026-08-10
2017ACM CCSPeer reviewed
Practical Secure Aggregation for Privacy-Preserving Machine Learning
Keith Bonawitz, Vladimir Ivanov, Ben Kreuter, Antonio Marcedone, H. Brendan McMahan, Sarvar Patel, Daniel Ramage, Aaron Segal, and Karn Seth
L2–L3
Privacy / federated
privacy / federated learning · secure aggregation · federated learning · pairwise masking · secret sharing · client dropout · robustness / domain shift / privacy
Official artifacts: Version-of-record manuscript · DOI
Metadata checked 2026-08-20
2017NeurIPSPeer reviewed
Prototypical Networks for Few-shot Learning
Jake Snell, Kevin Swersky, and Richard Zemel
L2
Generative / representation Loss / objectives Transfer / adaptation / continual
generative / representation learning · prototypical networks · few-shot learning · metric learning · loss / objective design
Metadata checked 2026-07-28
2017arXivTechnical report
Proximal Policy Optimization Algorithms
John Schulman et al.
L2
AI / reinforcement learning
AI / reinforcement learning · PPO · policy optimization · actor critic
Metadata checked 2026-07-27
2017CVPRPeer reviewed
Pyramid Scene Parsing Network
Hengshuang Zhao et al.
L2
Detection / segmentation
detection / segmentation · PSPNet · semantic segmentation · multiscale context
Metadata checked 2026-07-27
2017CVPRPeer reviewed
Quo Vadis, Action Recognition? A New Model and the Kinetics Dataset
Joao Carreira and Andrew Zisserman
L2
Video / 3D / embodied
video / 3D / embodied vision · I3D · action recognition · Kinetics
Metadata checked 2026-07-27
2017ICLRPeer reviewed
Semi-Supervised Classification with Graph Convolutional Networks
Thomas N. Kipf and Max Welling
L2
Neural networks Graph / geometric ML
neural networks · graph convolutional network · message passing · semi-supervised learning · graph / geometric learning
Metadata checked 2026-07-27
2017NeurIPSPeer reviewed
Simple and Scalable Predictive Uncertainty Estimation using Deep Ensembles
Balaji Lakshminarayanan, Alexander Pritzel, and Charles Blundell
L2
Evaluation / reproducibility Vision evaluation
evaluation / reproducibility · deep ensembles · uncertainty · calibration · vision evaluation / robustness
Metadata checked 2026-07-27
2017ICLRPeer reviewed
Understanding Deep Learning Requires Rethinking Generalization
Chiyuan Zhang et al.
L2–L3
Neural networks Evaluation / reproducibility
neural networks · evaluation / reproducibility · generalization · memorization · random labels
Metadata checked 2026-07-27
2017ICCVPeer reviewed
Unpaired Image-to-Image Translation Using Cycle-Consistent Adversarial Networks
Jun-Yan Zhu, Taesung Park, Phillip Isola, and Alexei A. Efros
L3
Generative vision Restoration / inverse problems
generative vision · generative / virtual pathology · CycleGAN · unpaired image translation · cycle consistency · adversarial learning
Official artifacts: Official implementation
Metadata checked 2026-08-10
2017ICMLPeer reviewed
Wasserstein Generative Adversarial Networks
Martin Arjovsky, Soumith Chintala, and Léon Bottou
L2–L3
Generative / representation
generative / representation learning · WGAN · Wasserstein distance · adversarial learning
Metadata checked 2026-07-27
2016KDDPeer reviewed
“Why Should I Trust You?”: Explaining the Predictions of Any Classifier
Marco Tulio Ribeiro, Sameer Singh, and Carlos Guestrin
L2–L3
Interpretability
interpretability / attribution · LIME · local surrogate explanations · model agnostic interpretability
Official artifacts: Official code
Metadata checked 2026-08-08
2016ICMLPeer reviewed
Asynchronous Methods for Deep Reinforcement Learning
Volodymyr Mnih et al.
L2
AI / reinforcement learning
AI / reinforcement learning · A3C · actor critic · asynchronous learning
Metadata checked 2026-07-28
2016ACM CCSPeer reviewed
Deep Learning with Differential Privacy
Martín Abadi et al.
L2–L3
Privacy / federated Neural networks
privacy / federated learning · neural networks · DP-SGD · privacy accounting · gradient clipping · robustness / domain shift / privacy
Official artifacts: Google differential privacy
Metadata checked 2026-08-08
2016CVPRPeer reviewed
Deep Residual Learning for Image Recognition
Kaiming He, Xiangyu Zhang, Shaoqing Ren, and Jian Sun
L1
Computer vision Neural networks Image / recognition
computer vision · foundation encoder · CNN · residual networks · neural networks · image recognition
Metadata checked 2026-07-27
2016JMLRPeer reviewed
Domain-Adversarial Training of Neural Networks
Yaroslav Ganin et al.
L2
Generative / representation Transfer / adaptation / continual
generative / representation learning · domain adaptation · gradient reversal · invariance · robustness / domain shift / privacy
Metadata checked 2026-07-27
2016arXivPreprint
Gaussian Error Linear Units (GELUs)
Dan Hendrycks and Kevin Gimpel
L2–L3
Neural networks
neural networks · GELU · activation function · stochastic regularization · smooth nonlinearity
Metadata checked 2026-08-20
2016CVPRPeer reviewed
Generation and Comprehension of Unambiguous Object Descriptions
Junhua Mao et al.
L1
Grounding / open vocabulary
grounding / open vocabulary · referring expressions · disambiguation · comprehension
Metadata checked 2026-07-28
2016arXivPreprint
Layer Normalization
Jimmy Lei Ba, Jamie Ryan Kiros, and Geoffrey E. Hinton
L1–L2
Neural networks
neural networks · layer normalization · normalization
Metadata checked 2026-07-27
2016ECCVPeer reviewed
Learning Without Forgetting
Zhizhong Li and Derek Hoiem
L2–L3
Peer-reviewed ECCV 2016 proceedings paper; the canonical locator is the open author manuscript.
Knowledge distillation Transfer / adaptation / continual Visual representations
knowledge distillation / model transfer · visual representation learning · Learning without Forgetting · continual learning · old-task distillation · new-task data · catastrophic forgetting
Official artifacts: Official project · Official code
Metadata checked 2026-08-23
2016NAACL-HLTPeer reviewed
Neural Architectures for Named Entity Recognition
Guillaume Lample, Miguel Ballesteros, Sandeep Subramanian, Kazuya Kawakami, and Chris Dyer
L1–L2
Text / sequence models NLP tasks / evaluation Structure / extraction
text / sequence modeling · NLP tasks / evaluation · named entity recognition · BiLSTM-CRF · structured prediction · linguistic structure / information extraction
Official artifacts: Code
Metadata checked 2026-07-28
2016ACLPeer reviewed
Neural Machine Translation of Rare Words with Subword Units
Rico Sennrich, Barry Haddow, and Alexandra Birch
L1
Text / sequence models Generation / MT evaluation
text / sequence modeling · BPE · subword tokenization · machine translation · generation / translation evaluation
Metadata checked 2026-07-27
2016ICMLPeer reviewed
Pixel Recurrent Neural Networks
Aäron van den Oord, Nal Kalchbrenner, and Koray Kavukcuoglu
L2
Generative / representation Generative vision
generative / representation learning · PixelRNN · PixelCNN · autoregressive modeling · generative vision
Metadata checked 2026-07-27
2016EMNLPPeer reviewed
Sequence-Level Knowledge Distillation
Yoon Kim and Alexander M. Rush
L2–L3
Knowledge distillation Generation / MT evaluation
knowledge distillation / model transfer · generation / translation evaluation · sequence-level distillation · neural machine translation · beam search
Official artifacts: Official seq2seq implementation · Official mobile demonstration
Metadata checked 2026-08-10
2016ECCVPeer reviewed
SPICE: Semantic Propositional Image Caption Evaluation
Peter Anderson, Basura Fernando, Mark Johnson, and Stephen Gould
L2–L3
Generation / MT evaluation Captioning / VQA V&L evaluation / safety
generation / translation evaluation · captioning / VQA · vision–language evaluation / safety · SPICE · image captioning · scene graph · semantic propositions · caption metric
Official artifacts: Author manuscript · Authors' implementation
Metadata checked 2026-08-20
2016EMNLPPeer reviewed
SQuAD: 100,000+ Questions for Machine Comprehension of Text
Pranav Rajpurkar et al.
L1–L2
NLP tasks / evaluation
NLP tasks / evaluation · SQuAD · question answering · benchmark
Metadata checked 2026-07-27
2016CVPRPeer reviewed
Structure-from-Motion Revisited
Johannes L. Schönberger and Jan-Michael Frahm
L2–L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · COLMAP · structure from motion · bundle adjustment · 3D / spatial vision
Metadata checked 2026-07-27
2016ICLRPeer reviewed
Unifying Distillation and Privileged Information
David Lopez-Paz, Léon Bottou, Bernhard Schölkopf, and Vladimir Vapnik
L3
ICLR 2016 conference paper; canonical locator is the complete arXiv manuscript.
Knowledge distillation Transfer / adaptation / continual
knowledge distillation / model transfer · generalized distillation · privileged information · teacher student · knowledge transfer
Metadata checked 2026-08-30
2016LRECPeer reviewed
Universal Dependencies v1: A Multilingual Treebank Collection
Joakim Nivre et al.
L2–L3
Structure / extraction Multilingual / low resource
linguistic structure / information extraction · multilingual / low-resource NLP · Universal Dependencies · treebanks · dependency syntax
Official artifacts: Universal Dependencies
Metadata checked 2026-08-08
2016ECCVPeer reviewed
Unsupervised Learning of Visual Representations by Solving Jigsaw Puzzles
Mehdi Noroozi and Paolo Favaro
Research
Visual representations
self-supervised learning · pretext task · jigsaw · visual representation learning
Metadata checked 2026-07-27
20163DVPeer reviewed
V-Net: Fully Convolutional Neural Networks for Volumetric Medical Image Segmentation
Fausto Milletari, Nassir Navab, and Seyed-Ahmad Ahmadi
L2–L3
Loss / objectives Detection / segmentation
loss / objective design · detection / segmentation · V-Net · soft Dice loss · volumetric segmentation
Official artifacts: Official code
Metadata checked 2026-08-10
2016KDDPeer reviewed
XGBoost: A Scalable Tree Boosting System
Tianqi Chen and Carlos Guestrin
L1–L2
Classical ML
classical machine learning · gradient boosting · trees · systems
Metadata checked 2026-07-27
2016CVPRPeer reviewed
You Only Look Once: Unified, Real-Time Object Detection
Joseph Redmon et al.
L1–L2
Detection / segmentation
detection / segmentation · YOLO · one-stage detection · real-time vision
Metadata checked 2026-07-27
2015ICLRPeer reviewed
Adam: A Method for Stochastic Optimization
Diederik P. Kingma and Jimmy Ba
L1–L2
ML / deep learning Neural networks
deep learning · optimization · neural networks
Metadata checked 2026-07-27
2015ICMLPeer reviewed
Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
Sergey Ioffe and Christian Szegedy
L1–L2
ML / deep learning Neural networks
deep learning · normalization · optimization · neural networks
Metadata checked 2026-07-27
2015CVPRPeer reviewed
CIDEr: Consensus-Based Image Description Evaluation
Ramakrishna Vedantam, C. Lawrence Zitnick, and Devi Parikh
L2–L3
Generation / MT evaluation Captioning / VQA V&L evaluation / safety
generation / translation evaluation · captioning / VQA · vision–language evaluation / safety · CIDEr · image captioning · consensus evaluation · TF-IDF · caption metric
Official artifacts: Official CVF PDF
Metadata checked 2026-08-20
2015NIPS Deep Learning and Representation Learning WorkshopTechnical report
Distilling the Knowledge in a Neural Network
Geoffrey Hinton, Oriol Vinyals, and Jeff Dean
L1–L2
Generative / representation Knowledge distillation
generative / representation learning · knowledge distillation · teacher–student learning · compression · knowledge distillation / model transfer
Metadata checked 2026-07-27
2015ICLRPeer reviewed
Explaining and Harnessing Adversarial Examples
Ian J. Goodfellow, Jonathon Shlens, and Christian Szegedy
L2–L3
Vision evaluation
vision evaluation / robustness · FGSM · adversarial examples · adversarial training · local linearity
Metadata checked 2026-08-10
2015CVPRPeer reviewed
FaceNet: A Unified Embedding for Face Recognition and Clustering
Florian Schroff, Dmitry Kalenichenko, and James Philbin
L2–L3
Loss / objectives Visual representations
loss / objective design · visual representation learning · triplet loss · metric learning · online triplet mining · face recognition · embedding
Metadata checked 2026-08-10
2015ICCVPeer reviewed
Fast R-CNN
Ross Girshick
L1–L2
Detection / segmentation
detection / segmentation · Fast R-CNN · object detection · RoI pooling
Metadata checked 2026-07-27
2015NeurIPSPeer reviewed
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks
Shaoqing Ren, Kaiming He, Ross Girshick, and Jian Sun
L2
Computer vision Detection / segmentation
computer vision · object detection · detection / segmentation
Metadata checked 2026-07-27
2015ICLRPeer reviewed
FitNets: Hints for Thin Deep Nets
Adriana Romero, Nicolas Ballas, Samira Ebrahimi Kahou, Antoine Chassang, Carlo Gatta, and Yoshua Bengio
L2–L3
Knowledge distillation Visual representations
knowledge distillation / model transfer · visual representation learning · FitNets · feature distillation · teacher hints · student compression
Official artifacts: Official code
Metadata checked 2026-08-10
2015CVPRPeer reviewed
Fully Convolutional Networks for Semantic Segmentation
Jonathan Long, Evan Shelhamer, and Trevor Darrell
L1–L2
Computer vision Detection / segmentation
computer vision · segmentation · dense prediction · detection / segmentation
Metadata checked 2026-07-27
2015CVPRPeer reviewed
Going Deeper with Convolutions
Christian Szegedy et al.
L1–L2
Image / recognition
image recognition · GoogLeNet · Inception · multiscale features
Metadata checked 2026-07-27
2015NaturePeer reviewed
Human-level Control through Deep Reinforcement Learning
Volodymyr Mnih et al.
L2
AI / reinforcement learning
AI / reinforcement learning · DQN · experience replay · deep reinforcement learning
Metadata checked 2026-07-27
2015ICCVPeer reviewed
Learning Spatiotemporal Features with 3D Convolutional Networks
Du Tran et al.
L1–L2
Video / 3D / embodied
video / 3D / embodied vision · C3D · 3D convolution · video representation
Metadata checked 2026-07-27
2015ICLRPeer reviewed
Neural Machine Translation by Jointly Learning to Align and Translate
Dzmitry Bahdanau, Kyunghyun Cho, and Yoshua Bengio
L1–L2
NLP Text / sequence models Generation / MT evaluation
NLP · attention · machine translation · text / sequence modeling · generation / translation evaluation
Metadata checked 2026-07-27
2015IEEE Transactions on RoboticsPeer reviewed
ORB-SLAM: A Versatile and Accurate Monocular SLAM System
Raúl Mur-Artal, J. M. M. Montiel, and Juan D. Tardós
L2–L3
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · 3D / spatial vision · ORB-SLAM · monocular SLAM · bundle adjustment · loop closure
Official artifacts: Official project page · Official implementation · arXiv record
Metadata checked 2026-08-20
2015CVPRPeer reviewed
Show and Tell: A Neural Image Caption Generator
Oriol Vinyals et al.
L1
Captioning / VQA
captioning / VQA · image captioning · encoder-decoder
Metadata checked 2026-07-27
2015ICMLPeer reviewed
Show, Attend and Tell: Neural Image Caption Generation with Visual Attention
Kelvin Xu et al.
L1–L2
Captioning / VQA
captioning / VQA · image captioning · visual attention
Metadata checked 2026-07-27
2015ICMLPeer reviewed
Trust Region Policy Optimization
John Schulman et al.
L2
AI / reinforcement learning
AI / reinforcement learning · TRPO · policy optimization · trust region
Metadata checked 2026-07-27
2015MICCAIPeer reviewed
U-Net: Convolutional Networks for Biomedical Image Segmentation
Olaf Ronneberger, Philipp Fischer, and Thomas Brox
L1–L2
Computer vision Detection / segmentation
computer vision · cell / tissue modeling · segmentation · biomedical imaging · detection / segmentation
Metadata checked 2026-07-27
2015ICCVPeer reviewed
Unsupervised Visual Representation Learning by Context Prediction
Carl Doersch, Abhinav Gupta, and Alexei A. Efros
Research
Visual representations
self-supervised learning · pretext task · spatial context · visual representation learning
Metadata checked 2026-07-27
2015ICLRPeer reviewed
Very Deep Convolutional Networks for Large-Scale Image Recognition
Karen Simonyan and Andrew Zisserman
L1
Image / recognition
image recognition · VGG · CNN · depth
Metadata checked 2026-07-27
2015ICCVPeer reviewed
VQA: Visual Question Answering
Stanislaw Antol et al.
L1
Captioning / VQA V&L evaluation / safety
captioning / VQA · vision–language evaluation / safety · visual question answering · dataset · human answers
Metadata checked 2026-07-27
2014ICLRPeer reviewed
Auto-Encoding Variational Bayes
Diederik P. Kingma and Max Welling
L2
Generative / representation Generative vision
generative / representation learning · variational autoencoder · variational inference · latent variables · generative vision
Metadata checked 2026-07-27
2014NeurIPSPeer reviewed
Do Deep Nets Really Need to Be Deep?
Lei Jimmy Ba and Rich Caruana
L2–L3
Peer-reviewed NeurIPS 2014 proceedings paper.
Knowledge distillation Neural networks Visual representations
knowledge distillation / model transfer · neural networks · visual representation learning · function matching · logit regression · shallow student · teacher ensemble · model compression
Official artifacts: Author manuscript
Metadata checked 2026-08-23
2014JMLRPeer reviewed
Dropout: A Simple Way to Prevent Neural Networks from Overfitting
Nitish Srivastava et al.
L1
ML / deep learning Neural networks
deep learning · regularization · neural networks
Metadata checked 2026-07-27
2014NeurIPSPeer reviewed
Generative Adversarial Nets
Ian Goodfellow et al.
L2
Generative / representation Generative vision
generative / representation learning · GAN · adversarial learning · implicit density · generative vision · generative / virtual pathology
Metadata checked 2026-07-27
2014EMNLPPeer reviewed
GloVe: Global Vectors for Word Representation
Jeffrey Pennington, Richard Socher, and Christopher D. Manning
L1
Text / sequence models
text / sequence modeling · GloVe · word embeddings · co-occurrence
Metadata checked 2026-07-27
2014NeurIPSPeer reviewed
How Transferable Are Features in Deep Neural Networks?
Jason Yosinski, Jeff Clune, Yoshua Bengio, and Hod Lipson
L2
Generative / representation Transfer / adaptation / continual
generative / representation learning · transfer learning · feature reuse · representation learning
Metadata checked 2026-07-27
2014ECCVPeer reviewed
Learning a Deep Convolutional Network for Image Super-Resolution
Chao Dong et al.
L1–L2
Generative vision Restoration / inverse problems
generative vision · SRCNN · super-resolution · restoration
Metadata checked 2026-07-27
2014CVPRPeer reviewed
Rich Feature Hierarchies for Accurate Object Detection and Semantic Segmentation
Ross Girshick et al.
L1–L2
Detection / segmentation
detection / segmentation · R-CNN · object detection · region proposals
Metadata checked 2026-07-27
2014NeurIPSPeer reviewed
Sequence to Sequence Learning with Neural Networks
Ilya Sutskever, Oriol Vinyals, and Quoc V. Le
L1
NLP Text / sequence models
NLP · encoder–decoder · text / sequence modeling
Metadata checked 2026-07-27
2014NeurIPSPeer reviewed
Two-Stream Convolutional Networks for Action Recognition in Videos
Karen Simonyan and Andrew Zisserman
L1–L2
Video / 3D / embodied
video / 3D / embodied vision · two-stream network · action recognition · optical flow
Metadata checked 2026-07-27
2013NeurIPSPeer reviewed
DeViSE: A Deep Visual-Semantic Embedding Model
Andrea Frome et al.
L1
Image–text representation Loss / objectives
image–text representation · visual-semantic embedding · zero-shot learning · ranking loss · loss / objective design
Metadata checked 2026-07-28
2013arXivPreprint
Efficient Estimation of Word Representations in Vector Space
Tomas Mikolov et al.
L1
NLP Text / sequence models
NLP · embeddings · text / sequence modeling
Metadata checked 2026-07-27
2013IEEE GlobalSIPPeer reviewed
Plug-and-Play Priors for Model Based Reconstruction
Singanallur V. Venkatakrishnan, Charles A. Bouman, and Brendt Wohlberg
L2–L3
Restoration / inverse problems
plug-and-play priors · ADMM · denoiser prior · tomographic reconstruction
Official artifacts: Author-hosted paper · Authors’ Plug-and-Play software and reading hub
Metadata checked 2026-08-20
2012NeurIPSPeer reviewed
ImageNet Classification with Deep Convolutional Neural Networks
Alex Krizhevsky, Ilya Sutskever, and Geoffrey E. Hinton
L1
Image / recognition Neural networks
image recognition · neural networks · AlexNet · CNN · ImageNet
Metadata checked 2026-07-27
2011JMLRPeer reviewed
Adaptive Subgradient Methods for Online Learning and Stochastic Optimization
John Duchi, Elad Hazan, and Yoram Singer
L2–L3
Neural networks
neural networks · AdaGrad · adaptive optimization · online learning · stochastic optimization
Metadata checked 2026-08-20
2011AISTATSPeer reviewed
Deep Sparse Rectifier Neural Networks
Xavier Glorot, Antoine Bordes, and Yoshua Bengio
L1
Neural networks
neural networks · ReLU · activation functions · sparsity
Metadata checked 2026-07-27
2010JMLRPeer reviewed
Composite Binary Losses
Mark D. Reid and Robert C. Williamson
L3
Loss / objectives
loss / objective design · proper losses · composite losses · class probability estimation
Metadata checked 2026-08-10
2010AISTATSPeer reviewed
Understanding the Difficulty of Training Deep Feedforward Neural Networks
Xavier Glorot and Yoshua Bengio
L1–L2
Neural networks
neural networks · initialization · gradient propagation
Metadata checked 2026-07-27
2009CVPRPeer reviewed
ImageNet: A Large-Scale Hierarchical Image Database
Jia Deng et al.
L1
Image / recognition
image recognition · ImageNet · dataset · benchmark
Metadata checked 2026-07-27
2008JMLRPeer reviewed
Visualizing Data using t-SNE
Laurens van der Maaten and Geoffrey Hinton
L2
Classical ML
classical machine learning · t-SNE · visualization · manifold learning
Metadata checked 2026-07-27
2006ECMLPeer reviewed
Bandit Based Monte-Carlo Planning
Levente Kocsis and Csaba Szepesvári
L2
AI / reinforcement learning
AI / reinforcement learning · Monte Carlo tree search · planning · upper confidence bounds
Metadata checked 2026-07-28
2006TCCPeer reviewed
Calibrating Noise to Sensitivity in Private Data Analysis
Cynthia Dwork, Frank McSherry, Kobbi Nissim, and Adam Smith
L2–L3
Privacy / federated
privacy / federated learning · differential privacy · global sensitivity · Laplace mechanism
Metadata checked 2026-08-08
2006CVPRPeer reviewed
Dimensionality Reduction by Learning an Invariant Mapping
Raia Hadsell, Sumit Chopra, and Yann LeCun
L2–L3
Loss / objectives Visual representations
loss / objective design · visual representation learning · contrastive loss · metric learning · invariance
Metadata checked 2026-08-10
2006KDDPeer reviewed
Model Compression
Cristian Buciluǎ, Rich Caruana, and Alexandru Niculescu-Mizil
L2–L3
Peer-reviewed KDD 2006 proceedings paper, pages 535–541.
Knowledge distillation Classical ML
knowledge distillation / model transfer · classical machine learning · model compression · ensemble transfer · pseudo-labelled transfer set · fast student · distillation precursor
Metadata checked 2026-08-23
2006ACM SIGGRAPHPeer reviewed
Photo Tourism: Exploring Photo Collections in 3D
Noah Snavely, Steven M. Seitz, and Richard Szeliski
L2
Video / 3D / embodied 3D / spatial vision
video / 3D / embodied vision · structure from motion · photo collections · 3D reconstruction · 3D / spatial vision
Metadata checked 2026-07-27
2006JMLRPeer reviewed
Statistical Comparisons of Classifiers over Multiple Data Sets
Janez Demšar
L2
Evaluation / reproducibility
evaluation / reproducibility · classifier comparison · multiple datasets · statistical testing
Metadata checked 2026-07-27
2005JMLRPeer reviewed
Estimation of Non-Normalized Statistical Models by Score Matching
Aapo Hyvärinen
L2–L3
Generative / representation
generative / representation learning · score matching · unnormalized density · energy-based model
Metadata checked 2026-07-27
2005CVPRPeer reviewed
Histograms of Oriented Gradients for Human Detection
Navneet Dalal and Bill Triggs
L1–L2
Image / recognition
image recognition · HOG · local descriptors · detection
Metadata checked 2026-07-27
2004International Journal of Computer VisionPeer reviewed
Distinctive Image Features from Scale-Invariant Keypoints
David G. Lowe
L1–L2
Image / recognition
image recognition · SIFT · local descriptors · matching
Metadata checked 2026-07-27
2004Text Summarization Branches OutPeer reviewed
ROUGE: A Package for Automatic Evaluation of Summaries
Chin-Yew Lin
L1
NLP tasks / evaluation Generation / MT evaluation
NLP tasks / evaluation · ROUGE · summarization · generation evaluation · generation / translation evaluation
Metadata checked 2026-07-28
2003JMLRPeer reviewed
A Neural Probabilistic Language Model
Yoshua Bengio, Réjean Ducharme, Pascal Vincent, and Christian Jauvin
L1–L2
Text / sequence models
text / sequence modeling · neural language model · distributed representation
Metadata checked 2026-07-27
2002ACLPeer reviewed
BLEU: a Method for Automatic Evaluation of Machine Translation
Kishore Papineni et al.
L1
NLP tasks / evaluation Generation / MT evaluation
NLP tasks / evaluation · BLEU · machine translation · evaluation · generation / translation evaluation
Metadata checked 2026-07-27
2002ACM Transactions on Information SystemsPeer reviewed
Cumulated Gain-Based Evaluation of IR Techniques
Kalervo Järvelin and Jaana Kekäläinen
L2–L3
NLP tasks / evaluation RAG / tools / agents Multimodal retrieval V&L evaluation / safety
NLP tasks / evaluation · RAG / tools / agents · multimodal embedding / retrieval · vision–language evaluation / safety · nDCG · graded relevance · ranking evaluation · information retrieval
Metadata checked 2026-08-20
2001ICMLPeer reviewed
Conditional Random Fields: Probabilistic Models for Segmenting and Labeling Sequence Data
John D. Lafferty, Andrew McCallum, and Fernando C. N. Pereira
L2–L3
Text / sequence models Structure / extraction
text / sequence modeling · linguistic structure / information extraction · conditional random field · sequence labeling · structured prediction · maximum entropy Markov model
Metadata checked 2026-08-20
2001The Annals of StatisticsPeer reviewed
Greedy Function Approximation: A Gradient Boosting Machine
Jerome H. Friedman
L2
Classical ML
classical machine learning · gradient boosting · functional gradient descent · ensembles
Metadata checked 2026-07-28
2001Machine LearningPeer reviewed
Random Forests
Leo Breiman
L1–L2
Classical ML
classical machine learning · random forest · bagging · decision trees
Metadata checked 2026-07-27
1999NeurIPSPeer reviewed
Policy Gradient Methods for Reinforcement Learning with Function Approximation
Richard S. Sutton, David A. McAllester, Satinder P. Singh, and Yishay Mansour
L2–L3
AI / reinforcement learning
AI / reinforcement learning · policy gradient · actor–critic · function approximation
Metadata checked 2026-07-27
1998Proceedings of the IEEEPeer reviewed
Gradient-Based Learning Applied to Document Recognition
Yann LeCun, Léon Bottou, Yoshua Bengio, and Patrick Haffner
L1
Neural networks
neural networks · LeNet · convolutional neural network · document recognition
Metadata checked 2026-07-27
1997Journal of Computer and System SciencesPeer reviewed
A Decision-Theoretic Generalization of On-Line Learning and an Application to Boosting
Yoav Freund and Robert E. Schapire
L1–L2
Classical ML
classical machine learning · AdaBoost · boosting · ensemble learning
Metadata checked 2026-07-28
1997Neural ComputationPeer reviewed
Long Short-Term Memory
Sepp Hochreiter and Jürgen Schmidhuber
L1–L2
Text / sequence models Neural networks
text / sequence modeling · neural networks · LSTM · recurrent neural network · gated memory
Metadata checked 2026-07-27
1996Journal of the Royal Statistical Society: Series BPeer reviewed
Regression Shrinkage and Selection via the Lasso
Robert Tibshirani
L1–L2
Classical ML
classical machine learning · L1 regularization · sparsity · linear regression
Metadata checked 2026-07-27
1995BiometrikaPeer reviewed
Causal diagrams for empirical research
Judea Pearl
L2–L3
Causal inference
causal inference · causal graphs · do-calculus · identification
Metadata checked 2026-08-08
1995Machine LearningPeer reviewed
Support-Vector Networks
Corinna Cortes and Vladimir Vapnik
L1–L2
Classical ML Loss / objectives
classical machine learning · support vector machine · kernel method · margin · loss / objective design
Metadata checked 2026-07-27
1994TREC-3Technical report
Okapi at TREC-3
Stephen E. Robertson, Steve Walker, Susan Jones, Micheline Hancock-Beaulieu, and Mike Gatford
L2–L3
System-description paper in NIST Special Publication 500-225 (TREC-3 proceedings); TREC participation and proceedings inclusion are not represented here as peer review.
NLP tasks / evaluation RAG / tools / agents
NLP tasks / evaluation · RAG / tools / agents · Okapi BM25 · probabilistic retrieval · term frequency saturation · document length normalization
Official artifacts: NIST proceedings record
Metadata checked 2026-08-20
1992Machine LearningPeer reviewed
Q-learning
Christopher J. C. H. Watkins and Peter Dayan
L2
AI / reinforcement learning
AI / reinforcement learning · Q-learning · temporal difference · control
Metadata checked 2026-07-27
1990ICMLPeer reviewed
Integrated Architectures for Learning, Planning, and Reacting Based on Approximating Dynamic Programming
Richard S. Sutton
L2
AI / reinforcement learning
AI / reinforcement learning · Dyna · model-based reinforcement learning · planning
Metadata checked 2026-07-27
1989Mathematics of Control, Signals and SystemsPeer reviewed
Approximation by Superpositions of a Sigmoidal Function
George Cybenko
L2–L3
Neural networks
neural networks · universal approximation · feedforward network · theory
Metadata checked 2026-07-27
1986IEEE TPAMIPeer reviewed
A Computational Approach to Edge Detection
John Canny
L1–L2
Image / recognition
image recognition · edge detection · image gradients · classical vision
Metadata checked 2026-07-27
1986Artificial IntelligencePeer reviewed
Fusion, Propagation, and Structuring in Belief Networks
Judea Pearl
L2
AI / reinforcement learning
AI / reinforcement learning · belief networks · probabilistic inference · graphical models
Metadata checked 2026-07-28
1986NaturePeer reviewed
Learning representations by back-propagating errors
David E. Rumelhart, Geoffrey E. Hinton, and Ronald J. Williams
L1
ML / deep learning Neural networks
deep learning · backpropagation · optimization · neural networks
Metadata checked 2026-07-27
1984Communications of the ACMPeer reviewed
A Theory of the Learnable
Leslie G. Valiant
L2–L3
Classical ML
classical machine learning · PAC learning · sample complexity · computational learning theory
Metadata checked 2026-07-27
1983BiometrikaPeer reviewed
The central role of the propensity score in observational studies for causal effects
Paul R. Rosenbaum and Donald B. Rubin
L2–L3
Causal inference
causal inference · propensity score · balancing score · strong ignorability · matching · subclassification
Official artifacts: Official article record
Metadata checked 2026-08-20
1982IEEE Transactions on Information TheoryPeer reviewed
Least Squares Quantization in PCM
Stuart P. Lloyd
L1–L2
Classical ML
classical machine learning · k-means · vector quantization · clustering
Metadata checked 2026-07-28
1981Communications of the ACMPeer reviewed
Random Sample Consensus: A Paradigm for Model Fitting with Applications to Image Analysis and Automated Cartography
Martin A. Fischler and Robert C. Bolles
L1–L2
Image / recognition
image recognition · RANSAC · robust estimation · geometry
Metadata checked 2026-07-27
1979The Annals of StatisticsPeer reviewed
Bootstrap Methods: Another Look at the Jackknife
Bradley Efron
L2
Evaluation / reproducibility
evaluation / reproducibility · bootstrap · resampling · uncertainty
Metadata checked 2026-07-27
1977Artificial IntelligencePeer reviewed
Consistency in Networks of Relations
Alan K. Mackworth
L2
AI / reinforcement learning
AI / reinforcement learning · constraint satisfaction · consistency · search
Metadata checked 2026-07-28
1977Journal of the Royal Statistical Society: Series BPeer reviewed
Maximum Likelihood from Incomplete Data via the EM Algorithm
Arthur P. Dempster, Nan M. Laird, and Donald B. Rubin
L2
Math / statistics Classical ML
mathematics / statistics · classical machine learning · expectation maximization · latent variables · maximum likelihood
Metadata checked 2026-07-27
1974Journal of Educational PsychologyPeer reviewed
Estimating causal effects of treatments in randomized and nonrandomized studies
Donald B. Rubin
L2–L3
Causal inference
causal inference · potential outcomes · average causal effect · randomization · matching · observational studies
Official artifacts: Official ETS precursor report
Metadata checked 2026-08-20
1972Journal of DocumentationPeer reviewed
A Statistical Interpretation of Term Specificity and Its Application in Retrieval
Karen Spärck Jones
L2–L3
Text / sequence models NLP tasks / evaluation RAG / tools / agents
text / sequence modeling · NLP tasks / evaluation · RAG / tools / agents · inverse document frequency · term specificity · information retrieval · term weighting
Metadata checked 2026-08-20
1968IEEE Transactions on Systems Science and CyberneticsPeer reviewed
A Formal Basis for the Heuristic Determination of Minimum Cost Paths
Peter E. Hart, Nils J. Nilsson, and Bertram Raphael
L1–L2
AI / reinforcement learning
AI / reinforcement learning · A* · heuristic search · optimal search
Metadata checked 2026-07-27
1958Psychological ReviewPeer reviewed
The Perceptron: A Probabilistic Model for Information Storage and Organization in the Brain
Frank Rosenblatt
L1–L2
Neural networks
neural networks · perceptron · linear classification · history
Metadata checked 2026-07-27
1951The Annals of Mathematical StatisticsPeer reviewed
A Stochastic Approximation Method
Herbert Robbins and Sutton Monro
L2
Math / statistics
mathematics / statistics · stochastic approximation · optimization
Metadata checked 2026-07-27
1948Bell System Technical JournalPeer reviewed
A Mathematical Theory of Communication
Claude E. Shannon
L2
Math / statistics
mathematics / statistics · entropy · information theory
Metadata checked 2026-07-27
1943Bulletin of Mathematical BiophysicsPeer reviewed
A Logical Calculus of the Ideas Immanent in Nervous Activity
Warren S. McCulloch and Walter Pitts
L1–L2
Neural networks
neural networks · threshold unit · neural computation · history
Metadata checked 2026-07-27
1901The London, Edinburgh, and Dublin Philosophical Magazine and Journal of SciencePeer reviewed
On Lines and Planes of Closest Fit to Systems of Points in Space
Karl Pearson
L1–L2
Math / statistics
mathematics / statistics · principal component analysis · linear algebra · dimensionality reduction
Metadata checked 2026-07-27

How to use it

Open a detailed chapter and understand its task and method taxonomy.
Filter this registry to the same route; compare adjacent papers by year, evidence state, and official artifacts.
Add only the papers you plan to study to My Library.
Record the claim, mechanism, data, evaluation, limitation, and one falsifying test.
Recheck the linked primary source before citing publication status, model access, reported performance, or license.

A paper can support more than one route. Transformer, CLIP, latent diffusion, and retrieval papers therefore keep one canonical catalog identity with multiple semantic memberships. My Library deduplicates that identity rather than creating parallel copies.

Current versus comprehensive

“Comprehensive” here means that every maintained scientific branch has prerequisites, foundational papers, major transitions, current endpoints, experiments, and an update policy. The Daily arXiv Radar provides breadth and recency. A radar record enters this trusted registry only after its identity, publication state, scientific contribution, route fit, primary links, and limitations are checked.

Recent preprints and technical reports are labelled as such. They are useful for following active research, but their claims remain provisional until the relevant evidence is independently tested.