List your company

Hallucination Detection on HaluEval GPT outputs (test)

0.582AUROCP(True)
0.513360.531180.5490.56682Jun 1, 2026

Evaluation results

MethodLinks
P(True)Scorer Category=WB2026.060.582
Non-Cont.Scorer Category=BB2026.060.579
Sem. Sets Conf.Scorer Category=BB2026.060.579
Mean Token Neg.Scorer Category=WB2026.060.579
GPT-4o judgeScorer Category=Judge2026.060.579
Prob. MarginScorer Category=WB2026.060.576
Semantic Neg.Scorer Category=BB2026.060.575
EntailmentScorer Category=BB2026.060.572
Cosine Sim.Scorer Category=BB2026.060.568
Sonnet judgeScorer Category=Judge2026.060.567
Sequence Prob.Scorer Category=WB2026.060.562
Gemini judgeScorer Category=Judge2026.060.562
Min Token Prob.Scorer Category=WB2026.060.528
Llama judgeScorer Category=Judge2026.060.518
Exact MatchScorer Category=BB2026.060.516