Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →brycewang-stanford/cvpr-supplementary
Use when deciding what belongs in a CVPR supplementary upload versus the 8-page body, covering the one-week-later supplement deadline, video and qualitative-result norms in computer vision, anonymous code packaging, the no-external-links rule, and keeping decision-critical evidence out of material reviewers may skip.
Passed 20 of 22 test cases with the skill, and 11 of 22 without it.
and 7 more in the full report
9 fixed11 passed either way2 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 9 test cases the skill fixed. Same model, same prompt, with and without the skill.
We restructured our supplementary PDF right before submission, dropping section E and merging it into D. What audit step must we perform on the main paper body before compiling?
| Checked | Without skill | With skill |
|---|---|---|
| Instructs searching the main paper body for 'Suppl' references. | Passed | Passed |
| Requires verifying that every body reference resolves to an existing supplement section. | Missed | Passed |
The instructions your model reads when the skill loads.
CVPR gives supplementary material its own deadline — November 20 in the 2026 cycle, one week after the November 13 paper deadline — and vision reviewers open supplements more than reviewers at most venues, because qualitative evidence lives there. That makes the supplement a designed deliverable, not an attic. Facts below are the 2026 cycle as read on 2026-07-08.
One rule sorts everything: anything a reviewer needs to judge the claims must be in the 8-page body. The supplement is legitimately for depth on things the body already establishes. A useful decision table:
| Content | Body or supplement? | Reasoning |
|---|---|---|
| Headline quantitative results | Body | The claim itself |
| Core ablation (the one that justifies the design) | Body | Reviewers rank papers on it |
| Full ablation grids, extra datasets | Supplement | Depth on an established point |
| Method details needed to believe the numbers | Body (compressed) | "See supplement" for correctness reads as evasion |
| Derivations, proofs, extra math | Supplement | Vision reviewers spot-check |
| Qualitative result grids beyond page 8 | Supplement | Expected and widely read |
| Failure cases | Both: one in body, catalog in supplement | Honesty signal reviewers reward |
| Video results (tracking, generation, 3D, robotics) | Supplement | The only sanctioned channel — external video links are banned |
| Code + configs | Supplement | See cvpr-artifact-evaluation |
For temporal or 3D work, the video is the result, and since external links are prohibited, the uploaded file is your one shot:
README.txt) mapping each clip to the paper section it supports.test cases passed, vs 11 / 22 without
Published skills whose descriptions closely match this one, with their headline benchmark lift.
More skills from the same source and nearby in the registry, with their headline benchmark lift.