Writing study: ANTS: Adaptive Negative Textual Space Shaping for OOD Detection via Test-Time MLLM Understanding and Reasoning

CVPR overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← CVPR 2026 oral writing corpus

Wenjie Zhu, Yabin Zhang, Xin Jin, Wenjun Zeng, Lei Zhang · Official CVPR 2026 oral record

Observed writing choices, not acceptance rules. This study examines the inspected manuscript version. It does not establish why the paper was accepted or that its presentation choices should be copied unchanged.

Figure, table and section ledger

Editorial contribution type: Architecture & representation (primary).

Layout extraction, not a required paper template. Counts are distinct caption labels; a multi-panel figure counts once. Section spans are estimated from heading positions, including floats and whitespace. Method and limitations.

Canonical PDF

11 physical pages

Before references / appendix

≈ 8.05 page-equivalents

Main figures / tables

7 / 6

Attached appendix figures / tables

0 / 0

Attached appendix span

Separate file recorded below; not in this PDF total

Before references / appendix: ≈ 73.18%References / intervening material: ≈ 26.82%Estimated layout envelopes, not text/artwork area. Unresolved appendix boundaries remain within the undivided post-body material.

Extraction notes: No attached appendix boundary detected. Zero means no recognized caption, not independently verified absence.

Section-by-section lengths and source pages

The title/abstract/front matter precede the first numbered heading. A span ends at the next detected top-level heading, not at the last sentence; end-page whitespace is included. References and statements are listed separately. These are not measurements of text-only or artwork area.

Top-level heading	Starts on PDF page	Estimated span	Share of entire PDF
1. Introduction	1	≈ 1.45 pp.	13.18%
2. Related Work	2	≈ 0.5 pp.	4.55%
3. Preliminary	3	≈ 0.77 pp.	7%
4. Methodology	4	≈ 1.8 pp.	16.36%
5. Experiments	5	≈ 2.86 pp.	26%
6. Conclusion and Future Work	8	≈ 0.27 pp.	2.45%
References	9	≈ 2.95 pp.	26.82%

All detected figure and table locations

Main text

• Figure 1 · p. 1
• Figure 2 · p. 2
• Figure 3 · p. 3
• Figure 4 · p. 4
• Figure 5 · p. 4
• Figure 6 · p. 5
• Table 1 · p. 6
• Table 2 · p. 7
• Table 3 · p. 7
• Table 4 · p. 7
• Figure 7 · p. 8
• Table 5 · p. 8
• Table 6 · p. 8

Attached appendix

No recognized caption labels.

Separate source files—excluded from canonical-PDF totals

A separate author version may repeat main text and references. Whole-file lengths and counts are not extra appendix material and are never added to the canonical PDF totals. Read the study's source scope for which sections received substantive reading versus layout inspection.

Additional source 1

Separate PDF (or ZIP containing it): whole-file layout of 8 physical pages; 4 detected figures and 7 tables. These may include overlapping main-text material.

Located appendix only: starts on physical p.1; ≈ 7.93 page-equivalents, 4 figure labels and 7 table labels. This interval is distinct from the whole file.

Detected heading	PDF page	Estimated span
A. More Results	1	≈ 0.81 pp.
B. Case Study	1	≈ 5.17 pp.
References	7	≈ 1.95 pp.

• Table 1 · p. 1
• Table 2 · p. 1
• Figure 1 · p. 1
• Table 3 · p. 2
• Table 4 · p. 2
• Table 5 · p. 2
• Table 6 · p. 3
• Table 7 · p. 3
• Figure 2 · p. 4
• Figure 3 · p. 5
• Figure 4 · p. 6

Argument and organization

1. Problem and contribution

   Maps three shortcomings of fixed negative labels to image understanding, near-OOD similarity and unknown test conditions.

2. Argument sequence

   Introduces the two textual spaces separately before the adaptive score that combines them.

3. Evidence and verification

   The main comparison separates training-required from zero-shot methods, while a second table distinguishes near-OOD and far-OOD outcomes.

Problem, claim and contribution

Maps three shortcomings of fixed negative labels to image understanding, near-OOD similarity and unknown test conditions.

Introduction, pp. 1–2 · Main paper · Additional source 2

Argument and section order

Introduces the two textual spaces separately before the adaptive score that combines them.

Sections 4.2–4.4, pp. 4–5 · Main paper · Additional source 2

Figures, captions and visual explanation

Figure 2 contrasts ordinary MLLM reasoning with its role in OOD detection using matching simple pipeline rows; later prompt examples make generated labels concrete.

Figures 2, 4–5, pp. 2 and 4 · Main paper · Additional source 2

Tables, comparisons and evidence

The main comparison separates training-required from zero-shot methods, while a second table distinguishes near-OOD and far-OOD outcomes.

Tables 1–2, pp. 6–7 · Main paper · Additional source 2

Appendix and reproducibility

Full benchmark results are supplemented by confidence-score cases separately for ID, far-OOD and near-OOD, so the two regimes remain distinct.

Supplement Figures 2–4, pp. 4–6 · Main paper · Additional source 2

Clarity and writing conventions

Algorithm 1 presents the online order from incoming batch through cache mining, text generation and scoring beside the benchmark protocol.

Algorithm 1, p. 6 · Main paper · Additional source 2

What to reuse—and what not to infer

Transferable lesson: Explain adaptive methods with the time order of available information, then evaluate the regimes that motivate each branch.

Boundary: A feature projection is illustrative, not a proof of semantic coverage; online historical-test access must remain explicit in comparisons.

Apply this lesson using the argument and evidence worksheet.

Inspection record and sources

Checked

2026-09-26

Reading status

Writing-focused manuscript scan + selected visuals inspected

Main text

Writing-oriented scan throughout Sections 1–6, PDF pp. 1–8; references pp. 9–11: abstract/framing, section sequence, figure/table captions, evaluation organization and conclusions; not an exhaustive technical verification.

Appendix / supplement

Writing-oriented supplement scan: Supplement Appendices A–B, pp. 1–8: full OOD tables, noisy-cache robustness and ID/near-/far-OOD case studies. Video/code assets were not executed or assessed.

Visually inspected PDF pages

2, 6

This is a writing-focused scan of the manuscript and the stated supplementary scope: framing, organization, visual explanation, evidence presentation and wording. It is not a line-by-line proof audit, reproduction of results or exhaustive scientific review.

• Official CVF open-access CVPR 2026 accepted-paper PDF · 11 pages
• Official CVF supplementary PDF · 8 pages