StateBench: Testing how language models track state changes over long documents
Comments
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Ranked by a public rubric, not by votes: what a comment adds counts, not who wrote it. How comments are ranked · Comment rules
Summary
StateBench, released by Allen Institute for AI researchers, is a benchmark that tests how well language models track evolving entity states across long documents, using four difficulty levels from immutable assignment to general finite swaps.