AISTATS Writing Style
Use this when revising the main paper. AISTATS papers need compact statements of why a
statistical or ML contribution matters and enough detail for technical validation.
Revision rules
- Put the AI/statistics contribution in the first page: problem, gap, method or theorem,
evidence, and why existing methods are insufficient.
- Make assumptions explicit. AISTATS reviewers are sensitive to hidden distributional,
asymptotic, optimization, or data-generating assumptions.
- Pair every major claim with either proof structure, simulation design, empirical table, or
reproducibility detail.
- Use the 8-page body for core logic; move long derivations, extra simulations, and extended
ablations to the appendix without making the main paper unintelligible.
- Avoid overclaiming empirical wins when differences are small or variance is unreported.
- Maintain double-blind style in self-citations, prior code references, acknowledgements,
funding, and artifact descriptions.
Theorem-presentation discipline
- Number assumptions and cite them by label inside every theorem statement; AISTATS readers
audit assumption flows the way software reviewers audit imports.
- Give each major theorem a proof sketch in the body; the appendix proof supports the sketch
but never replaces it.
- Keep problem-dependent constants visible or explicitly deferred; silently absorbing
dimension or condition numbers into O-notation is a standing statistician complaint.
- Define notation once in one location — two-column pages punish redundant redefinition.
- Phrase remarks after theorems as interpretation, not as extra unproved claims; AISTATS
reviewers treat every declarative sentence near a theorem as something they may verify.
- When a result is conjectured or only empirically supported, label it so; mixing proved and
observed statements in one paragraph is a credibility leak at this venue.
Sentence-level rewrites
| Draft pattern |
AISTATS-safe rewrite |
| "Our method significantly outperforms..." |
"reduces mean error by X (SE Y) over Z seeded runs" |
| "Under mild conditions..." |
"Under Assumptions 1-3 (boundedness, ...)" |
| "It is easy to see that..." |
"By Lemma 2 and the triangle inequality..." |
| "Achieves state-of-the-art..." |
Claim scoped to the regimes actually tested |
Vignette: compressing into eight two-column pages
A draft with three theorems, six figures, and a sprawling related-work section: keep all
theorem statements, one sketch each, and the two decision-critical figures (the rate plot
and the coverage plot); compress related work into contribution contrasts; move secondary
lemmas and four figures to the appendix with explicit forward references. The test of a good
cut: a reviewer should reconstruct the whole argument without ever opening the supplement.
Output format
[Writing diagnosis] clear / under-justified / overclaimed / overloaded
[First-page fix] <new framing>
[Claim discipline] <claim -> proof/experiment/limitation>
[Compression cuts] <move/delete/merge>
[Anonymity edits] <phrases to rewrite>
Source: brycewang-stanford/Awesome-Journal-Skills → AISTATS-Skills/skills/aistats-writing-style/SKILL.md
1---2name: aistats-writing-style3description: Use when revising an AISTATS paper for concise AI-statistics framing, theorem-and-experiment clarity, 8-page two-column compression, double-blind wording, reproducibility clarity, statistically careful claims, and assumption-labeling discipline that survives statistician reviewers.4---567# AISTATS Writing Style89Use this when revising the main paper. AISTATS papers need compact statements of why a10statistical or ML contribution matters and enough detail for technical validation.1112## Revision rules1314- Put the AI/statistics contribution in the first page: problem, gap, method or theorem,15 evidence, and why existing methods are insufficient.16- Make assumptions explicit. AISTATS reviewers are sensitive to hidden distributional,17 asymptotic, optimization, or data-generating assumptions.18- Pair every major claim with either proof structure, simulation design, empirical table, or19 reproducibility detail.20- Use the 8-page body for core logic; move long derivations, extra simulations, and extended21 ablations to the appendix without making the main paper unintelligible.22- Avoid overclaiming empirical wins when differences are small or variance is unreported.23- Maintain double-blind style in self-citations, prior code references, acknowledgements,24 funding, and artifact descriptions.2526## Theorem-presentation discipline2728- Number assumptions and cite them by label inside every theorem statement; AISTATS readers29 audit assumption flows the way software reviewers audit imports.30- Give each major theorem a proof sketch in the body; the appendix proof supports the sketch31 but never replaces it.32- Keep problem-dependent constants visible or explicitly deferred; silently absorbing33 dimension or condition numbers into O-notation is a standing statistician complaint.34- Define notation once in one location — two-column pages punish redundant redefinition.35- Phrase remarks after theorems as interpretation, not as extra unproved claims; AISTATS36 reviewers treat every declarative sentence near a theorem as something they may verify.37- When a result is conjectured or only empirically supported, label it so; mixing proved and38 observed statements in one paragraph is a credibility leak at this venue.3940## Sentence-level rewrites4142| Draft pattern | AISTATS-safe rewrite |43|---|---|44| "Our method significantly outperforms..." | "reduces mean error by X (SE Y) over Z seeded runs" |45| "Under mild conditions..." | "Under Assumptions 1-3 (boundedness, ...)" |46| "It is easy to see that..." | "By Lemma 2 and the triangle inequality..." |47| "Achieves state-of-the-art..." | Claim scoped to the regimes actually tested |4849## Vignette: compressing into eight two-column pages5051A draft with three theorems, six figures, and a sprawling related-work section: keep all52theorem statements, one sketch each, and the two decision-critical figures (the rate plot53and the coverage plot); compress related work into contribution contrasts; move secondary54lemmas and four figures to the appendix with explicit forward references. The test of a good55cut: a reviewer should reconstruct the whole argument without ever opening the supplement.5657## Output format5859```text60[Writing diagnosis] clear / under-justified / overclaimed / overloaded61[First-page fix] <new framing>62[Claim discipline] <claim -> proof/experiment/limitation>63[Compression cuts] <move/delete/merge>64[Anonymity edits] <phrases to rewrite>65```6667---6869**Source:** [`brycewang-stanford/Awesome-Journal-Skills`](https://github.com/brycewang-stanford/Awesome-Journal-Skills) → `AISTATS-Skills/skills/aistats-writing-style/SKILL.md`
thedixitjain