Benchmark
ALCE
A benchmark for evaluating the citation attributes of language models, combining natural language inference to judge whether a document chunk semantically entails the answer assertion.
Timeline (last 90 days)
Oct 17
CSR (Citation Survival Rate) requires each citation to pass two checks: a support check that uses FActScore-style atomic claim decomposition plus an ALCE-style entailment model to verify whether the cited block supports the claim, and a provenance check that relies on tool-call logs to confirm the URL actually appeared in the content the Agent retrieved.
Unverified50%