Claude Skills by brycewang-stanford

Use when positioning a COLM paper inside the fastest-moving literature in ML — triaging arXiv-heavy citations, handling concurrent work fairly, citing model and dataset artifacts correctly, distinguishing COLM's three-edition archive from adjacent venues' archives, and keeping self-citation double-blind safe.

Use when hardening a COLM paper's reproducibility story — pinning open-weight checkpoints and tokenizers, handling API-model drift and deprecation honestly, versioning evaluation harnesses and prompts, disclosing compute, and writing availability statements that distinguish what is releasable from what is not.

Use when reasoning about how COLM reviews a paper — the OpenReview pipeline from late-March submission through the May review release, the May-June rebuttal window, July decisions, reciprocal-reviewing obligations, the LLM-use rules for reviewers, and how a three-edition-old venue's norms differ from mature conferences.

Use when preparing or auditing a COLM submission on OpenReview — the late-March abstract and full-paper deadlines, the strict 9-page main text, double-blind rules banning acknowledgments and identity links, the Code of Ethics acknowledgment, LLM-usage disclosure, reciprocal-reviewer nomination, and pre-upload risk triage.

Use when deciding what goes into a COLM paper's appendices and supplementary material versus the strict 9-page main text — verbatim prompts, full evaluation configurations, per-task result tables, contamination analyses, human-evaluation protocols, and anonymized code/data packages that survive double-blind review.

Use when deciding whether language-model research belongs at COLM or should route to ACL/EMNLP, ICLR, NeurIPS, ICML, or a workshop — applying the object-of-study test, matching against COLM's CFP lanes (training, data, evaluation, inference, safety), and weighing the trade-offs of a young venue before writing begins.

Use when planning a COLM submission campaign across the calendar — working backward from the late-March abstract and paper deadlines through the May-June rebuttal, July decisions, August camera-ready, and October conference, coordinating co-authors, compute, and reciprocal-reviewing duties, and slotting COLM into a multi-venue LM-research pipeline.

Use when drafting or revising COLM paper prose — leading with a finding about language models rather than a leaderboard delta, scoping claims to tested models and scales, naming versions in text, keeping the 9-page main text self-sufficient, and matching the measured, analysis-forward voice of COLM's award lineage.

Use when deciding what evidence package a COLT (Conference on Learning Theory) paper needs, given that COLT runs no artifact-evaluation track or badges — the proof appendix is the artifact. Covers proof-verification passes, optional code companions for numerics, formalization aids, and post-acceptance release of scripts.

Use when drafting a COLT (Conference on Learning Theory) rebuttal after initial reviews, handling proof-correctness objections, tightness and novelty-of-technique challenges, the area-chair identity-reveal rule, and concession strategy for theorem-first papers where one unrepaired gap outweighs every polished answer.

Use when converting an accepted COLT (Conference on Learning Theory) paper into its PMLR proceedings version, covering de-anonymization, acknowledgements and funding restoration, appendix consolidation into the archival PDF, PMLR volume metadata and BibTeX hygiene, rebuttal-promised edits, and presentation logistics.

Use when deciding whether a COLT (Conference on Learning Theory) paper needs numerical content at all — COLT has no experiments requirement — and, when numerics genuinely help, designing small illustrative simulations that visualize a proved bound, a separation, or a phase transition without diluting the theory.

Use when positioning a COLT (Conference on Learning Theory) submission against prior bounds and models — building the known-versus-new comparison across COLT/ALT lineage, STOC/FOCS, NeurIPS-and-ICML theory tracks, statistics journals, and arXiv concurrency, while respecting anonymity and the parallel-submission rules.

Use when strengthening the reproducibility of a COLT (Conference on Learning Theory) paper, where reproducing means re-deriving — complete proofs, explicit assumptions, tracked constants, correctly invoked external results, self-contained notation — plus seeds and scripts for any numerical illustration the paper carries.

Use when reasoning about COLT (Conference on Learning Theory) peer review — the double-anonymous-with-informed-area-chair model, correctness-first evaluation by expert theorists, the rebuttal stage before decisions, single-track acceptance stakes, and how PMLR publication and the community's proof culture shape outcomes.

Use when auditing a COLT (Conference on Learning Theory) submission for CMT readiness, the 12-page PMLR-format main body, single-PDF assembly with unlimited appendix, double-anonymous formatting, the parallel-submission ban covering journals and proceedings venues, and the ordered final-week checks before the AoE deadline.

Use when structuring the appendix of a COLT (Conference on Learning Theory) submission — the unlimited-length proof appendix inside the single PDF — covering body/appendix splitting for a 12-page limit, theorem restatement discipline, lemma ordering, notation tables, and what referees expect to find where.

Use when deciding whether a result is COLT-shaped (Conference on Learning Theory) — a theorem-first learning-theory contribution — versus better routed to ALT, STOC/FOCS, NeurIPS/ICML/AISTATS, JMLR, or a statistics journal, and whether the right vehicle is a full paper or a COLT open-problem piece.

Use when planning a COLT (Conference on Learning Theory) submission calendar backward from the single AoE paper deadline — proof-completion and verification milestones, writing and appendix-assembly passes, CMT logistics, the rebuttal window, decision aftermath, and PMLR camera-ready with conference presentation in early summer.

Use when revising a COLT (Conference on Learning Theory) paper for theorem-first exposition — formal setup before results, informal-then-formal theorem statements, proof overviews that sell the technique, quantifier and constant hygiene, and fitting the argument's spine into 12 PMLR-formatted pages.

Use when packaging what stands behind a CSCW paper — systems, analysis pipelines, codebooks, instruments, datasets from real communities — for review-time scrutiny and post-acceptance release, where community-data ethics constrain release more than any badge checklist.

Use when a CSCW paper draws Revise and Resubmit or Revise for External Review — triaging reviewer requests, running the weeks-long revision as a project, and writing the point-by-point response letter that the same reviewers will read before they reread anything else.

Use when a CSCW paper is accepted and enters PACMHCI production — executing minor-change conditions, deanonymizing text and artifacts, ACM journal e-rights and TAPS steps, issue assignment, and the separate track of presenting at the next CSCW conference.

Use when designing or auditing the empirical work behind a CSCW paper — interview and ethnographic rigor, trace and log analysis, surveys, deployments, and mixed methods — matching each method's own validity standard and the ethics of studying real communities.

Use when positioning a CSCW paper's related work — engaging the venue's own concept lineage, the PACMHCI-era literature, neighboring HCI venues, and the social sciences CSCW borrows from, with era-correct citations and anonymity-safe self-references for journal-model review.

Use when strengthening the transparency of a CSCW paper — auditable qualitative analysis trails, documented trace pipelines, codebooks and instruments, and honest data-availability statements when community and participant data cannot ethically be shared.

Use when interpreting where a CSCW paper stands in review — the PACMHCI journal model's decision vocabulary (Revise for External Review, Revise and Resubmit, Conditional Accept), same-reviewer re-review, the rolling editorial board, and what each outcome statistically implied in the last cycle.

Use when readying a CSCW manuscript for upload — confirming the current pathway (rolling via Manuscript Central as of mid-2026), the single-column ACM template, the word-length scrutiny band, anonymization of author and community identifiers, ethics statements, and desk-risk triage.

Use when assembling supplementary materials for a CSCW submission — appendices, codebooks, interview guides, survey instruments, analysis code, and data extracts — deciding what accompanies a journal-model paper, anonymized end to end and consistent across R&R rounds.

Use when deciding whether a project belongs at CSCW — is the group, community, or organization the real unit of analysis? — and running the routing decision against CHI, ICWSM, DIS, GROUP, TOCHI, and social-science journals before any drafting or platform account setup begins.

Use when planning a CSCW submission timeline end to end — the PACMHCI journal model, the retired fixed cycles versus the rolling 2027+ pathway, Revise-and-Resubmit rounds that span months, and back-planning so an acceptance lands before the conference-year presentation cutoff.

Use when drafting or revising CSCW prose — group-level framing, concepts a field can reuse, participant voice handled with care, theory that earns its keep, and length discipline under the 5,000-12,000-word scrutiny band where contribution is weighed against every page.

Use when packaging code, models, datasets, or demo videos for a CVPR paper at either review time or release time, covering anonymous supplement packaging under the external-link ban, the dataset-release-by-camera-ready rule, model-weight and license decisions, and making a vision artifact runnable by a skeptical stranger.

Use when planning or drafting the CVPR one-page rebuttal after reviews are released, covering the official rebuttal template, the ban on new contributions and external links, triaging multiple reviews at 16k-submission scale, choosing which numbers fit in one page, and writing for the AC who reads the discussion.

Use when preparing the final version of an accepted CVPR paper, covering de-anonymization, the dataset-release-by-camera-ready obligation, CVF open access versus the IEEE Xplore version of record, oral/highlight/poster preparation at conference scale, and arXiv/preprint synchronization after acceptance.

Use when designing or auditing the experimental program of a CVPR paper, covering benchmark and baseline selection under matched-compute fairness, the ablation study reviewers treat as mandatory, qualitative and failure-case evidence, efficiency metrics tied to the Compute Reporting Form, and generalization tests beyond a single dataset.

Use when building the related-work section and positioning for a CVPR paper, covering literature that moves at arXiv speed, concurrent-work etiquette, the CVPR/ICCV/ECCV sibling triangle, double-blind self-citation, verifying that cited "CVPR papers" are actually CVPR papers, and defining deltas against the true nearest neighbors.

Use when hardening a CVPR paper's reproducibility story, covering the Compute Reporting Form's hardware and compute sections, training-recipe disclosure, benchmark protocol and split hygiene, seed and variance reporting for vision experiments, and closing the gaps reviewers probe at a benchmark-driven venue.

Use when reasoning about how CVPR review actually works at 16,000-submission scale, covering the OpenReview pipeline and timeline, reviewer-duty enforcement and desk rejects, the reviewer LLM ban, AC and discussion dynamics, oral/highlight/poster decision tiers, and calibrating expectations to a ~25% acceptance rate.

Use when auditing a CVPR submission before the OpenReview deadline, covering abstract registration and profile requirements, the 8-page limit including figures and tables, the Compute Reporting Form, anonymity and external-link bans, dual-submission attestations, reviewer-duty enrollment, and desk-reject triage.

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.

Use when deciding whether a project belongs at CVPR or should route elsewhere, covering what counts as a vision contribution at the field's flagship, fit tests for methods, datasets, and application papers, realistic odds at 25% acceptance and 16k submissions, and routing to ICCV, ECCV, WACV, 3DV, NeurIPS, or a journal.

Use when planning a CVPR submission campaign end to end, covering the November abstract/paper/supplement deadline chain, coauthor reviewer-duty scheduling, the January rebuttal sprint, February decisions, camera-ready and June conference logistics, and the resubmission ladder across ICCV, ECCV, and the next CVPR.

Use when writing or revising a CVPR paper's prose and figures, covering the 8-page budget where figures and tables count against the limit, the page-1 teaser figure convention, benchmark-table craft, claim calibration for a skeptical vision audience, and taking responsibility for tool-assisted text under CVPR's rules.

在为投向《计算机学报》(Chinese Journal of Computers, CJC) 的稿件准备代码与数据可用性材料时调用。本刊目前没有像国际计算机会议(如 ACM/USENIX artifact evaluation)那样独立的制品评审徽章制度，本技能讲清这一现状，并指导作者如何自愿地、规范地随长文提供可复现的代码、数据与实验脚本，如何在正文中声明可用性、如何托管到稳定仓库并给出访问方式，从而增强外审专家对计算机全学科实证结果的信任与可核验性。适用于让中文原创研究的支撑材料经得起三审推敲的场景。

在收到《计算机学报》(Chinese Journal of Computers, CJC) 外审意见、退修通知或多轮修回要求后，撰写审稿意见答复信与修回说明时调用。覆盖本刊三审制下"退修/大修/小修"的定位判断、逐条对应式回复的结构、修改稿差异标注、与主编终审衔接的答复语气、外审专家分歧的处理，以及重新送审情形下如何组织再审材料。适用于把一份计算机全学科中文长文从退修状态推进到录用的多轮沟通场景。

在稿件被《计算机学报》(Chinese Journal of Computers, CJC) 录用后处理定稿、校样与清样时调用。覆盖录用后的版权协议与著作权转让、按官方模板做最终排版、中英文摘要与作者简介的定稿核对、中图分类号与 DOI/基金信息的完善、参考文献 GB/T 7714 最终校订、校样(proof)阶段的勘误纪律与仅限编辑性改动的原则，以及版面费缴纳与出版排期确认。适用于把一篇已录用的计算机全学科中文长文安全送到正式出版的场景。

在为投向《计算机学报》(Chinese Journal of Computers, CJC) 的长文设计与呈现实验、评估或理论验证时调用。覆盖如何按贡献类型(算法/系统/理论/实证/学习类)匹配证据、如何设立评估问题与公平基线、如何选数据集与指标、如何做消融与显著性分析、如何报告均值方差与统计检验而非单点最优、如何防数据泄漏与评测污染、以及如何用图表在长文中充分而不冗余地呈现结果。适用于让计算机全学科中文原创研究的实验部分经受本刊三审专家对"数据充分真实、结论可信"的审查。

在为投向《计算机学报》(Chinese Journal of Computers, CJC) 的长文撰写相关工作与文献综述时调用。覆盖综合性期刊读者跨学科的特点下如何组织文献脉络、如何写"delta 优先"的定位段把本文与已有工作的差异讲清、如何兼顾国内外文献与经典—前沿的平衡、如何规范引用(GB/T 7714 与中文文献中英文对照)、如何避免过度自引与堆引、以及综述型稿件与研究论文中相关工作的不同写法。适用于让计算机全学科中文原创研究在三审中清晰确立创新性与学术坐标的场景。

在为投向《计算机学报》(Chinese Journal of Computers, CJC) 的实证或系统类长文构建可复现性与实验可重复保障时调用。覆盖实验环境与依赖固定、随机种子与统计稳定性、数据来源与预处理的可追溯记录、基线与超参数的公平呈现、机器学习/数据挖掘研究的数据泄漏与污染防范、复现脚本与论文表图的一一对应，以及在双盲外审下如何组织可核验又不暴露作者身份的复现材料。适用于让计算机全学科中文原创研究经受三审专家对结果可信度的审查。