Writing study: ViT^3: Unlocking Test-Time Training in Vision

CVPR overviewCorpus atlasBy contribution typeVisual galleryRules & templateWriting patternsPaper studies

← CVPR 2026 oral writing corpus

Dongchen Han, Yining Li, Tianyu Li, Zixuan Cao, Ziming Wang, Jun Song, Cheng Yu, Bo Zheng, Gao Huang · 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: Empirical analysis, diagnosis & auditing (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

≈ 8.11 page-equivalents

Main figures / tables

4 / 10

Attached appendix figures / tables

0 / 0

Attached appendix span

Separate file recorded below; not in this PDF total

Before references / appendix: ≈ 73.73%References / intervening material: ≈ 26.27%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	≈ 1.27 pp.	11.55%
2. Related Work	2	≈ 0.28 pp.	2.55%
3. Preliminaries	2	≈ 0.93 pp.	8.45%
4. Exploring Test-Time Training Designs	3	≈ 3.4 pp.	30.91%
5. ViT3: A Test-Time Training Architecture	7	≈ 1.59 pp.	14.45%
6. Conclusion	8	≈ 0.21 pp.	1.91%
Acknowledgements	9	≈ 0.06 pp.	0.55%
References	9	≈ 2.89 pp.	26.27%

All detected figure and table locations

Main text

• Figure 1 · p. 1
• Figure 2 · p. 3
• Table 1 · p. 4
• Table 2 · p. 4
• Table 3 · p. 4
• Table 4 · p. 5
• Figure 3 · p. 5
• Table 5 · p. 6
• Table 6 · p. 6
• Table 7 · p. 6
• Table 8 · p. 7
• Table 9 · p. 8
• Table 10 · p. 8
• Figure 4 · 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 3 physical pages; 0 detected figures and 3 tables. These may include overlapping main-text material.

Located appendix only: starts on physical p.1; ≈ 2.94 page-equivalents, 0 figure labels and 3 table labels. This interval is distinct from the whole file.

Detected heading	PDF page	Estimated span
Supplementary Material	1	≈ 0.02 pp.
1. Contribution and Limitation	1	≈ 0.2 pp.
2. Inner Training Loss	1	≈ 0.93 pp.
3. Model Architecture	2	≈ 0.19 pp.
References	2	≈ 1.6 pp.

• Table 1 · p. 2
• Table 2 · p. 2
• Table 3 · p. 2

Argument and organization

1. Problem and contribution

   Distinguishes promise of linear-complexity test-time training from the still-unsettled visual design choices.

2. Argument sequence

   Runs a design-space exploration before introducing ViT3, so inner-loss, batch and model choices precede the final model comparison.

3. Evidence and verification

   Tables 1–3 report parameter count, FLOPs or throughput alongside accuracy and explicitly mark diverged runs, retaining an important negative result.

Problem, claim and contribution

Distinguishes promise of linear-complexity test-time training from the still-unsettled visual design choices.

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

Argument and section order

Runs a design-space exploration before introducing ViT3, so inner-loss, batch and model choices precede the final model comparison.

Sections 4–5, pp. 3–8 · Main paper · Additional source 2

Figures, captions and visual explanation

Figure 2 puts the replacement block beside the familiar Transformer scaffold, making the changed computation easy to locate.

Figure 2, p. 3 · Main paper · Additional source 2

Tables, comparisons and evidence

Tables 1–3 report parameter count, FLOPs or throughput alongside accuracy and explicitly mark diverged runs, retaining an important negative result.

Tables 1–3, p. 4 · Main paper · Additional source 2

Appendix and reproducibility

Exact inner-loss expressions and tabulated architecture configurations are placed in the supplement along with contribution/limitation discussion.

Supplement Sections 1–3, pp. 1–3 · Main paper · Additional source 2

Clarity and writing conventions

Numbered insights are immediately followed by experiments or gradient arguments; remarks distinguish interpretation from measured settings.

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

What to reuse—and what not to infer

Transferable lesson: Build a design study around controlled questions and visible failures, then let the final architecture instantiate the resulting lessons.

Boundary: Empirical design insights are conditional on the backbones, training recipe and tasks studied, not universal rules for every TTT system.

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 Sections 1–3, pp. 1–3: contribution/limitations, explicit inner losses and architecture configurations. Video/code assets were not executed or assessed.

Visually inspected PDF pages

3, 4

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 · 3 pages