Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Deception Detection on InstrumentalEval (InstEval)
Loading...
0.867
AUROC
STATEWITNESS
0.16292
0.34571
0.5285
0.71129
Jun 16, 2026
AUROC
Updated 1mo ago
Evaluation Results
Method
Method
Links
AUROC
STATEWITNESS
Target Model=GPT-OSS-20B
2026.06
0.867
STATEWITNESS
Target Model=GEMMA4-26...
2026.06
0.823
LLM-as-a-Judge
Target Model=GPT-OSS-20B
2026.06
0.741
LLM-as-a-Judge
Target Model=GEMMA4-26...
2026.06
0.681
CMMS probe
Target Model=GPT-OSS-20B
2026.06
0.665
MMS probe
Target Model=GPT-OSS-20B
2026.06
0.663
LLM-as-a-Judge w/ CoT
Target Model=GEMMA4-26...
2026.06
0.636
CMMS probe
Target Model=GEMMA4-26...
2026.06
0.634
LLM-as-a-Judge w/ CoT
Target Model=GPT-OSS-20B
2026.06
0.622
Self-Evaluation
Target Model=GPT-OSS-20B
2026.06
0.607
LR probe
Target Model=GEMMA4-26...
2026.06
0.588
Self-Evaluation
Target Model=GEMMA4-26...
2026.06
0.558
LAT probe
Target Model=GEMMA4-26...
2026.06
0.506
MMS probe
Target Model=GEMMA4-26...
2026.06
0.5
LR probe
Target Model=GPT-OSS-20B
2026.06
0.33
LAT probe
Target Model=GPT-OSS-20B
2026.06
0.19
Feedback
Search any
task
Search any
task