Agent Skills
› brycewang-stanford/Awesome-Journal-Skills
› eacl-related-work
eacl-related-work
GitHub指导撰写EACL论文相关工作部分，强调对比论证而非罗列。要求优先引用ACL Anthology并正确标注Findings，公平处理并发预印本，进行近期文献新鲜度扫描，涵盖LREC-COLING等欧洲及多语言社区，避免引文错误。
Trigger Scenarios
撰写NLP会议论文的Related Work章节
需要对比分析最新NLP文献与竞品
检查EACL相关引文的准确性与时效性
Install
npx skills add brycewang-stanford/Awesome-Journal-Skills --skill eacl-related-work -g -y
SKILL.md
Frontmatter
{
"name": "eacl-related-work",
"description": "Use when positioning an EACL paper against fast-moving NLP literature, covering ACL-Anthology-first citation, correct Findings attribution, handling concurrent preprints fairly, freshness sweeps across recent *ACL rounds and European venues like LREC-COLING, and framing related work as a contrastive argument rather than a list."
}
EACL Related Work
Use this to build an EACL related-work section that positions the contribution rather than cataloguing the field. NLP literature moves fast and much of it is on arXiv before it is archival, so the section must be current, correctly attributed, and contrastive. Reopen the ACL Anthology and recent *ACL proceedings before finalizing.
Cite from the Anthology first
- The ACL Anthology is the canonical record for *ACL papers; cite the Anthology version (with its venue and year) rather than an arXiv preprint when both exist. This fixes venue, year, and page numbers correctly.
- For a paper that appeared in Findings, attribute it as Findings (e.g. "Findings of ACL: EACL 2026"), not as the main conference — misattributing Findings as main is a common, noticeable error.
Positioning, not listing
| Weak related work | Strong related work |
|---|---|
| "X did A. Y did B. Z did C." | "Unlike X, which assumes A, we ..." |
| A neutral literature dump | A contrast that sets up the gap |
| Only the famous references | The nearest competitors, named precisely |
| Silent on concurrent work | Concurrent preprints acknowledged fairly |
Handle concurrent preprints fairly
- If a closely related preprint appeared shortly before or during your cycle, acknowledge it and state what is independent or different about your work — since ARR has no anonymity period, reviewers may well know both.
- Do not claim novelty a reviewer can disprove with a two-week-old arXiv paper; scope the claim instead.
Freshness sweep
Before submitting, sweep for missed neighbours:
- recent ACL / EMNLP / NAACL / AACL / EACL proceedings in the Anthology
- Findings volumes (often missed)
- LREC-COLING for resources / multilingual / lower-resource work
- TACL and Computational Linguistics for journal-track results
- arXiv for the last ~3 months of concurrent preprints
Mind the European and multilingual neighbourhood
- EACL's community overlaps heavily with LREC-COLING (resources, multilingual, lower-resource languages) and EAMT (machine translation). A related-work section on a multilingual or resource paper that ignores these venues will read as under-read to EACL reviewers.
- Cite work on the specific languages you study, including regional and non-English-language publications where relevant.
Avoid sibling-venue misattribution
- The Anthology hosts every *ACL venue; confirm each citation's venue line before labeling it
"EACL." See the sibling-confusion guard in
../../resources/exemplars/library.md.
Output format
[Positioning] Contrastive / Listy
[Nearest competitors] <named, with the contrast to each>
[Attribution errors] <Findings-as-main, wrong venue, preprint-over-Anthology>
[Freshness gaps] <recent *ACL / LREC-COLING / preprints missed>
[Multilingual neighbourhood] <European/resource venues to add>
[Fix order] <what to revise>
Version History
- 9f86f09 Current 2026-07-19 15:14
