A hand-rolled index of AI papers that actually ship source code

What it does This repository is a community-curated set of Markdown tables mapping accepted papers from nine major AI venues—ACL, EMNLP, NAACL, COLING, AAAI, IJCAI, ICLR, NeurIPS, and ICML—to their publicly available source code. The lists aggregate links from DBLP, ACL Anthology, OpenReview, and PapersWithCode, organized by conference and year for quick browsing.

The interesting bit In an era of automated aggregators, this project relies on manual curation by a sizable contributor team. That human-in-the-loop approach is deliberately low-tech, but it serves as a filter: every entry is supposed to actually have runnable code attached, not just a promise of release.

Key highlights

• Covers nine venues spanning NLP, general AI, and core machine learning.
• Sourced from established paper databases including DBLP, ACL Anthology, and PapersWithCode.
• Maintained by roughly 18 named contributors; open to issues and pull requests.
• Version v0.1.0 with a “building” status, so coverage is still expanding.
• Several recent years are already populated (e.g., ACL 2019–2022, NeurIPS 2019–2021).

Caveats

• Many years—especially 2023–2025—are currently empty placeholders with no paper links.
• The README never clarifies whether every code-bearing paper is included or just a selection.
• One internal link (AAAI 2021) accidentally points to a differently named repository (NLP-Conferences-Code), hinting at some organizational drift.

Verdict Worth bookmarking if you need a quick, human-filtered starting point for reproducing results from specific conference years. If you need full coverage, advanced search, or automatic link validation, look elsewhere.