List your company

Hallucination Detection on LLaMa 1 (test)

0.894AUROCEigenTrack
0.73280.774650.81650.85835Jan 24, 2026

Evaluation results

MethodLinks
EigenTrackModel Scale=7B2026.010.894
LapEigvalsModel Scale=7B2026.010.871
EigenTrackModel Scale=3B2026.010.861
HaloScopeModel Scale=7B2026.010.861
EigenTrackModel Scale=1B2026.010.842
INSIDEModel Scale=3B2026.010.831
HaloScopeModel Scale=3B2026.010.827
HaloScopeModel Scale=1B2026.010.82
LapEigvalsModel Scale=3B2026.010.819
INSIDEModel Scale=7B2026.010.81
SelfCheckGPTModel Scale=7B2026.010.809
SelfCheckGPTModel Scale=3B2026.010.804
LapEigvalsModel Scale=1B2026.010.785
INSIDEModel Scale=1B2026.010.753
SelfCheckGPTModel Scale=1B2026.010.739