List your company

Hallucination detection on FinHalu

94.1PrecisionGPT-4o (teacher)
68.41275.08181.7588.419Apr 26, 2026

Evaluation results

MethodLinks
GPT-4o (teacher)role=teacher, retrieval-equalized=true2026.0494.195.995
FINGROUND (8B distilled)model_size=8B, type=distilled, retrieval-equalized=true2026.0492.790.291.4
Self-RAGretrieval-equalized=true2026.0481.277.179.1
CRAGretrieval-equalized=true2026.0480.674.977.6
HHEMretrieval-equalized=true2026.0478.973.876.3
FActScoreretrieval-equalized=true2026.0474.279.376.7
SelfCheckGPTretrieval-equalized=true2026.0469.476.572.8