Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› eacl-submission
eacl-submission
GitHub用于在EACL投稿前审计ARR周期提交，涵盖格式限制、匿名性、强制性局限章节及声明合规性，防止因单周期无回退机制导致的拒稿。
Trigger Scenarios
准备向ACL Rolling Review (ARR) 提交论文以参加 EACL 会议
检查论文是否符合 EACL/ARR 的特定格式和匿名要求
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill eacl-submission -g -y
SKILL.md
Frontmatter
{
"name": "eacl-submission",
"description": "Use when auditing an EACL submission before its single ACL Rolling Review cycle deadline, covering ACL-template compliance, the 8-page long \/ 4-page short content caps, the mandatory Limitations section, the Responsible NLP checklist, anonymization, preprint and resubmission declarations, and the no-fallback-cycle timing trap."
}
EACL Submission
Run this audit before uploading to the ACL Rolling Review (ARR) cycle that feeds your target
EACL edition. EACL papers are never submitted to the conference directly: they enter an ARR
cycle on OpenReview, collect reviews and a meta-review, and are later committed to EACL,
which decides main / Findings / reject. Reopen the current EACL CFP and aclrollingreview.org/cfp
before trusting any number below.
The one-cycle reality (EACL's defining trap)
- Confirm which single ARR cycle feeds your edition, and treat it as unrepeatable. For EACL 2027 the organizers named the August 3, 2026 cycle the only viable cycle — the conference sits early in the year, so there is no later cycle to slip into. Unlike ACL, a missed or wasted cycle here loses the whole edition.
- The ARR deadline is the hard one. Commitment is a separate, later step (October 11, 2026
for EACL 2027) — see
eacl-workflow. - With no fallback, a thin-but-submittable draft is a real dilemma: submitting burns the only cycle, skipping loses the year. Decide deliberately, not by deadline reflex.
Format gate
- Long papers: up to 8 content pages. Short papers: up to 4. Both get unlimited pages for references and appendices, plus unlimited space after the conclusion for the required Limitations section and an optional ethics statement.
- A missing Limitations section or overlength content is an explicit desk-reject condition under the ARR CFP — the most common self-inflicted rejection.
- Use the official ACL LaTeX/Word style files unmodified; margin, font, or spacing tampering is treated as a length violation and can be rejected without review.
- Appendices go after references in the same PDF; reviewers are not obliged to read them, so the 8 or 4 content pages must carry the whole argument.
Anonymity gate
- The submission must be fully anonymized: no names, affiliations, acknowledgements, or self-revealing "in our prior work we showed" citations.
- Since January 2024 there is no anonymity period: a non-anonymous preprint is allowed at any time. But declare preprint status at submission; choosing "no non-anonymous preprint" earns award eligibility and borderline priority — then posting one anyway before meta-review release is a desk-reject offense.
- Supplementary archives, repo links, and dataset pages must be anonymized too; links to personally identifying hosting are disallowed.
Declarations that gate the ARR form
| Form item | What ARR checks | Failure mode |
|---|---|---|
| Responsible NLP checklist | Completed, consistent with the PDF | Misleading answers → desk reject |
| Resubmission link | Prior ARR submission ID + change summary | Unacknowledged resubmission → desk reject |
| Preprint declaration | Matches what is actually on arXiv | Broken "no preprint" pledge → desk reject |
| Dual submission | Not under review anywhere archival | Violation → reject at ARR or EACL |
| AI assistance disclosure | Checklist item answered honestly | Undisclosed generative-AI use → policy breach |
Long vs short
A short paper is a genre, not a truncated long paper: one claim, four pages. A compressed
eight-page study reads as thin (see the fastText exemplar in
../../resources/exemplars/library.md). See eacl-topic-selection for the full choice.
Final-72-hours sequence for an NLP paper
- Rebuild the PDF from the untouched style files and re-run the page count on content pages only.
- Re-read Limitations: it must be substantive (languages, data, compute, scope), not a ritual paragraph.
- Fill the Responsible NLP checklist against the actual PDF — every "yes" needs a section number.
- Sweep the supplement archive for author names in paths, notebook metadata, git history, and license headers.
- Confirm the OpenReview title/abstract match the PDF and the preprint declaration matches arXiv reality.
- Submit hours early; ARR deadlines are AoE but OpenReview outages are not an accepted excuse, and there is no next cycle.
Frequently mishandled cases
- Assuming a fallback cycle exists. For a single-cycle edition it does not; plan as if this is the only attempt.
- Appendix used as overflow body. Content pages must carry the argument; "see Appendix B" for a core mechanism fails self-containedness.
- arXiv v2 posted mid-cycle after declaring no preprint: a desk-reject liability.
- Dataset "available on request": allowed, but checklist answers must then explain reproduction honestly rather than imply release.
Output format
[ARR readiness] Ready / Needs fixes / Not ready
[Cycle plan] <single viable cycle -> EACL edition; fallback = none>
[Format gate] <length/Limitations/style-file findings>
[Anonymity + declarations] <checklist/preprint/resubmission/dual-submission findings>
[Desk-reject risks] <ranked list>
[Fix order] <what to change before the one cycle deadline>
Version History
- 9f86f09 Current 2026-07-19 15:15
