List your company

Deception Detection on Insider Trading

100AUROCSelf-Evaluation
44.46458.88273.387.718Jun 16, 2026

Evaluation results

MethodLinks
Self-EvaluationTarget Model=GEMMA4-26B-A4B-IT2026.06100
STATEWITNESSTarget Model=GEMMA4-26B-A4B-IT2026.06100
LLM-as-a-Judge w/ CoTTarget Model=GEMMA4-26B-A4B-IT2026.0699.7
LLM-as-a-JudgeTarget Model=GEMMA4-26B-A4B-IT2026.0699.6
LAT probeTarget Model=GEMMA4-26B-A4B-IT2026.0698.6
STATEWITNESSTarget Model=GPT-OSS-20B2026.0698.2
LR probeTarget Model=GEMMA4-26B-A4B-IT2026.0690.6
LLM-as-a-Judge w/ CoTTarget Model=GPT-OSS-20B2026.0689.3
LLM-as-a-JudgeTarget Model=GPT-OSS-20B2026.0687.9
MMS probeTarget Model=GEMMA4-26B-A4B-IT2026.0682.7
Self-EvaluationTarget Model=GPT-OSS-20B2026.0673.2
CMMS probeTarget Model=GPT-OSS-20B2026.0667.2
CMMS probeTarget Model=GEMMA4-26B-A4B-IT2026.0658
LAT probeTarget Model=GPT-OSS-20B2026.0656.6
MMS probeTarget Model=GPT-OSS-20B2026.0655
LR probeTarget Model=GPT-OSS-20B2026.0646.6