ICLR Supplementary
Use this when deciding what belongs in the main PDF, appendix, supplementary files, or private
discussion-period links. The goal is to make extra evidence easy to find without hiding the paper's
core argument outside the main text.
Structure
- Keep central method, claims, and minimum evidence in the main text.
- Put derivations, extended ablations, robustness, extra qualitative examples, hyperparameters,
model cards, dataset cards, and ethics details in the appendix.
- Use supplementary files for code, data, large tables, logs, demos, and artifacts that do not fit
cleanly in the PDF.
- Include a one-page appendix map at the start when the appendix is long.
- Label discussion-period revisions explicitly so reviewers can find changes quickly.
Anonymity and access
- Remove author names, institutions, usernames, Git remotes, file paths, cloud buckets, and license
headers that identify the team.
- Avoid links that reveal visitors or owner accounts. Prefer anonymized repositories, static files,
or private OpenReview-visible links permitted by the current guide.
- Make supplementary filenames descriptive but neutral.
- If code cannot be released, state the legal or privacy constraint and provide a smaller
reproducibility substitute.
Main-text versus appendix placement
ICLR's single-track format means the main PDF carries the argument and appendices come after
references with generous room. The risk is burying decisive evidence where a time-pressed reviewer
never looks, while padding the appendix with material that proves nothing.
| Content |
Belongs in |
Why |
| Central claim and its key result |
Main text |
Reviewers may not open the appendix |
| Full proof / extended derivation |
Appendix |
Needed for rigor, not for the headline |
| Hyperparameters, model/dataset cards |
Appendix |
Reference material, not argument |
| Code, logs, large tables, checkpoints |
Supplementary ZIP |
Too big or interactive for the PDF |
| Decisive ablation a claim depends on |
Main text |
Moving it weakens the narrative |
Worked vignette
A generative-model paper hides its only fairness-of-comparison ablation in Appendix G, so a reviewer
concludes the headline gain is unsupported. The fix: promote a compact version into the main text
near the claim, leave the full grid in the appendix, and add an appendix map. During discussion the
authors upload a revision and label the changelog so the relocation is obvious.
Reviewer-pushback patterns
- "Key evidence is buried." Promote it; an appendix map is not a substitute for main-text placement.
- "Cannot tell what changed in the revision." Add a dated changelog at the top of the revised PDF.
- "Supplement leaks identity." Re-check file paths, license headers, and remotes before upload.
Output format
[Supplement plan] PDF appendix / ZIP / anonymous repo / private link / no supplement
[Main-text dependencies] <claims that must move back into main text>
[Reviewer navigation] <appendix map and file names>
[Anonymity risks] <paths, metadata, links, licenses>
[Discussion update note] <short changelog if revised>
Source: brycewang-stanford/Awesome-Journal-Skills → ICLR-Skills/skills/iclr-supplementary/SKILL.md
1---2name: iclr-supplementary3description: Use when organizing ICLR appendices, supplementary files, anonymous code/data, revised PDFs, private links, and discussion-period updates under OpenReview rules. Use when deciding what stays in the main text versus the appendix, how to label a discussion-period revision so reviewers find changes fast, or how to package anonymous artifacts that remain part of the permanent public record after acceptance.4---5
6
7# ICLR Supplementary
8
9Use this when deciding what belongs in the main PDF, appendix, supplementary files, or private
10discussion-period links. The goal is to make extra evidence easy to find without hiding the paper's
11core argument outside the main text.
12
13## Structure
14
15- Keep central method, claims, and minimum evidence in the main text.
16- Put derivations, extended ablations, robustness, extra qualitative examples, hyperparameters,
17 model cards, dataset cards, and ethics details in the appendix.
18- Use supplementary files for code, data, large tables, logs, demos, and artifacts that do not fit
19 cleanly in the PDF.
20- Include a one-page appendix map at the start when the appendix is long.
21- Label discussion-period revisions explicitly so reviewers can find changes quickly.
22
23## Anonymity and access
24
25- Remove author names, institutions, usernames, Git remotes, file paths, cloud buckets, and license
26 headers that identify the team.
27- Avoid links that reveal visitors or owner accounts. Prefer anonymized repositories, static files,
28 or private OpenReview-visible links permitted by the current guide.
29- Make supplementary filenames descriptive but neutral.
30- If code cannot be released, state the legal or privacy constraint and provide a smaller
31 reproducibility substitute.
32
33## Main-text versus appendix placement
34
35ICLR's single-track format means the main PDF carries the argument and appendices come after
36references with generous room. The risk is burying decisive evidence where a time-pressed reviewer
37never looks, while padding the appendix with material that proves nothing.
38
39| Content | Belongs in | Why |
40| --- | --- | --- |
41| Central claim and its key result | Main text | Reviewers may not open the appendix |
42| Full proof / extended derivation | Appendix | Needed for rigor, not for the headline |
43| Hyperparameters, model/dataset cards | Appendix | Reference material, not argument |
44| Code, logs, large tables, checkpoints | Supplementary ZIP | Too big or interactive for the PDF |
45| Decisive ablation a claim depends on | Main text | Moving it weakens the narrative |
46
47## Worked vignette
48
49A generative-model paper hides its only fairness-of-comparison ablation in Appendix G, so a reviewer
50concludes the headline gain is unsupported. The fix: promote a compact version into the main text
51near the claim, leave the full grid in the appendix, and add an appendix map. During discussion the
52authors upload a revision and label the changelog so the relocation is obvious.
53
54## Reviewer-pushback patterns
55
56- "Key evidence is buried." Promote it; an appendix map is not a substitute for main-text placement.
57- "Cannot tell what changed in the revision." Add a dated changelog at the top of the revised PDF.
58- "Supplement leaks identity." Re-check file paths, license headers, and remotes before upload.
59
60## Output format
61
62```text
63[Supplement plan] PDF appendix / ZIP / anonymous repo / private link / no supplement
64[Main-text dependencies] <claims that must move back into main text>
65[Reviewer navigation] <appendix map and file names>
66[Anonymity risks] <paths, metadata, links, licenses>
67[Discussion update note] <short changelog if revised>
68```
69
70---
71
72**Source:** [`brycewang-stanford/Awesome-Journal-Skills`](https://github.com/brycewang-stanford/Awesome-Journal-Skills) → `ICLR-Skills/skills/iclr-supplementary/SKILL.md`
thedixitjain