List your company
ResearchFormerly WizWand

Hallucination Detection on CoQA (Accuracy)

79.19AccuracyVal Loss
57.287662.973868.6674.3462May 25, 2026
Updated 5mo ago

Evaluation results

MethodLinks
Val LossModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0579.19
IDModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0579.19
FEPoIDModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0579.19
CurvatureModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0577.37
SNRModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0576.39
IDModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0573.56
FEPoIDModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0573.56
Val LossModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0573.29
CurvatureModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0571.9
SNRModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0568.59
RGNModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0567.86
RGNModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0564.21
RankMEModel=LlaMA-3.2-3B, Forward horizon (w)=7, Representation extraction (FST)=true2026.0562.65
RankMEModel=LlaMA-3.2-1B, Forward horizon (w)=3, Representation extraction (FST)=true2026.0558.13