AI Safety Literature Database
Two complementary databases of technical AI safety work — academic papers and community forum posts — each with structured metadata, summaries, and quality ratings.
Papers
3,700+ academic papers from arXiv with citation networks, methodology details, and reusable artifacts.
63k citation edges·1,100+ datasets·6k tags
Forum Posts
4,700+ synthesized posts from LessWrong, Alignment Forum, EA Forum, and technical blogs.
123k link edges·11k comments·5.5k tags
Both databases share the same 13 research area categories and are cross-referenced — forum posts link to papers and vice versa.