List your company

Hallucination Detection on TriviaQA GPT outputs (test)

0.675AUROCP(True)
0.568920.596460.6240.65154Jun 1, 2026

Evaluation results

MethodLinks
P(True)Scorer Category=WB2026.060.675
Non-Cont.Scorer Category=BB2026.060.653
Sonnet judgeScorer Category=Judge2026.060.651
EntailmentScorer Category=BB2026.060.64
Semantic Neg.Scorer Category=BB2026.060.639
Sem. Sets Conf.Scorer Category=BB2026.060.639
Min Token Prob.Scorer Category=WB2026.060.617
Sequence Prob.Scorer Category=WB2026.060.613
Mean Token Neg.Scorer Category=WB2026.060.607
Exact MatchScorer Category=BB2026.060.603
Cosine Sim.Scorer Category=BB2026.060.599
Prob. MarginScorer Category=WB2026.060.595
Llama judgeScorer Category=Judge2026.060.589
Gemini judgeScorer Category=Judge2026.060.582
GPT-4o judgeScorer Category=Judge2026.060.573