Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
FEVEROUSgithub.com · Low evidenceFEVEROUS is a dataset and benchmark for fact extraction and verification over both unstructured text and structured tables from Wikipedia. It contains 87,026 verified claims with evidence annotations and a baseline model that predicts evidence and verdicts for 18% of claims.
Parse Score