List your company

Hallucination Detection on XSum (AUROC)

0.73AUROCINTRYGUEmin-max
0.61560.64530.6750.7047Mar 23, 2026

Evaluation results

MethodLinks
INTRYGUEmin-maxBackbone=Mistral-7B2026.030.73
INTRYGUEmeanBackbone=Mistral-7B2026.030.7
LN-EntropyBackbone=Mistral-7B2026.030.69
ReDeEPBackbone=Qwen3-8B2026.030.69
INTRYGUEmin-maxBackbone=Qwen3-8B2026.030.68
MaxEntropyBackbone=Mistral-7B2026.030.67
Semantic EntropyBackbone=Qwen3-8B2026.030.67
INTRYGUEmin-maxBackbone=Llama-2-7B2026.030.66
INTRYGUEmin-maxBackbone=Llama-3.1-8B2026.030.66
EigenScoreBackbone=Llama-3.1-8B2026.030.66
INTRYGUEmin-maxBackbone=Llama-2-13B2026.030.63
INTRYGUEmin-maxBackbone=gemma-3-4b2026.030.62