eacl-experiments
Use when designing or auditing the empirical evidence for an EACL paper, covering tuned and LLM baselines, multilingual breadth matched to the claim, significance and variance floors, human-evaluation agreement, data-contamination controls, ablations, and error taxonomies, so that every stated result is measured rather than asserted.
Profile →npx skills add brycewang-stanford/Awesome-Journal-Skills --skill eacl-experiments --agent claude-code
Same command for any agent — swap --agent for codex, cursor, copilot.
Weekly change comes from our own snapshots, not the repository page — it measures attention, not adoption.
# EACL Experiments Use this to make an EACL paper's evidence hold up under NLP review. EACL rewards **well-scoped questions answered with careful controls** over leaderboard maximalism — its best papers include analyses and critiques, not only state-of-the-art systems (see `../../resources/exemplars/library.md`). Design the evidence to match the claim exactly, no broader. ## Baselines that make a comparison fair - Include a **tuned** baseline, not a strawman: an under-tuned competitor makes a win meaningless. State the search space for both your method and the baselines. - For LLM-based work, include the **obvious prompt/few-shot baseline** and report its prompts and decoding settings; a gain over an unreported baseline is not credible. ## Match breadth to the claim | Claim | Required breadth | |---|---| | "Works for language L" | Solid results on L, honestly scoped | | "Cross-lingual / multilingual" | Enough languages across resource levels; per-language results | | "General method" | Multiple tasks/datasets, not one convenient benchmark | | "Robust" | Stress tests / shifts, not just in-distribution | A multilingual claim backed by two high-resource languages is the classic EACL o
- Baselines that make a comparison fair
- Match breadth to the claim
- Significance and variance floor
- Contamination controls
- Human evaluation done properly
- Error analysis as a first-class result
- Audit checklist
- Output format
What does the eacl-experiments skill do?
Use when designing or auditing the empirical evidence for an EACL paper, covering tuned and LLM baselines, multilingual breadth matched to the claim, significance and variance floors, human-evaluation agreement, data-contamination controls, ablations, and error taxonomies, so that every stated result is measured rather than asserted.
How do I install it?
Run `npx skills add brycewang-stanford/Awesome-Journal-Skills --skill eacl-experiments --agent claude-code` — it drops the skill into your project so the agent can pick it up. Swap the --agent value for codex, cursor or copilot if you use one of those.
Where does this skill come from?
From brycewang-stanford/Awesome-Journal-Skills, a repository with 909 stars. We read it straight from the repository tree rather than a submitted listing, so what you see here is what is actually published.
Is a popular skill a good skill?
Not necessarily. Stars measure attention, not adoption — a repository can trend for a week and be abandoned. That is why we show the weekly change from our own snapshots next to the total, instead of a single flattering number.