{"schema_version":1,"total":1,"offset":0,"limit":100,"observations":[{"id":"public:8d7dac6e6ac868bc672d0a91","benchmark_id":"bullshitbench-v2::snapshot-2026-09-10","subject":{"source_id":"anthropic/claude-sonnet-5@reasoning=low","name":"anthropic/claude-sonnet-5@reasoning=low","model_id":"claude-sonnet-5::low","variant":null,"harness":null},"value":0.8,"unit":"fraction","basis":"measured","source":{"url":"https://raw.githubusercontent.com/petergpt/bullshit-benchmark/2678ac296fe6e234d391e0f4ab339f6a777ba2a3/data/v2/latest/leaderboard.csv","retrieved_at":"2026-09-15T01:07:23.764154+00:00","published_at":null,"file":"data/raw/benchmarks/daily-evidence/2026-09-15-bullshitbench/0c866538642eab944f67.gz","sha256":"0c866538642eab944f672eb1464857ad3f4cb2de89d484360ad6fd0765b8a390","locator":"csv; source row 12; anthropic/claude-sonnet-5@reasoning=low; field green_rate"},"protocol":"BullshitBench V2 (snapshot-2026-09-10); 100 nonsense prompts; three-judge panel (Claude Sonnet 4.6, GPT-5.2, Gemini 3.1 Pro Preview) grades clear pushback / partial challenge / accepted nonsense; value = green_rate over all attempts; one row per model x reasoning label; not a Composite input.; reasoning=low; avg_score=1.7172; score_2=80; nonsense_count=100; answered_count=99; refusal_count=1; error_count=0","comparison_key":null,"join_note":"Reviewed identity map 2026-09-16: label states model claude-sonnet-5 and setting low; exact catalog configuration claude-sonnet-5::low"}],"divergences":[],"cell":null,"coverage":{"model":null,"benchmark":{"total_models":924,"available":75,"unknown":849,"not_tested":0,"not_published":0,"source_unreachable":0,"contested":0,"measured":75,"self_reported":0,"observations":214,"unmatched_observations":139}},"collection":{"benchmark_id":"bullshitbench-v2::snapshot-2026-09-10","status":"collected","source_url":"https://raw.githubusercontent.com/petergpt/bullshit-benchmark/2678ac296fe6e234d391e0f4ab339f6a777ba2a3/data/v2/latest/leaderboard.csv","reason":"214 source results parsed; configurations remain separate; unmatched model identities are retained."},"coverage_note":"Denominators are catalog model configurations × versioned registry entries; capability_available excludes Efficiency (cost) boards. Basis counts count covered cells, not runs; measured and self_reported may overlap. Unmatched source identities remain in observations and never inflate catalog coverage."}