List your company

Truthfulness on TruthfulQA

81MA.PROBELLM
43.5653.286372.72Feb 13, 2026

Evaluation results

MethodLinks
PROBELLMTarget Model=olmo-3-7b-ins.2026.02811964
PROBELLMTarget Model=Claude-3.5-sonnet2026.02762454
PROBELLMTarget Model=devstral2026.02663447
PROBELLMTarget Model=Llama-3.1-8b-ins.2026.02623878
PROBELLMTarget Model=granite-4.02026.02584270
PROBELLMTarget Model=Deepseek-v3.22026.02574343
PROBELLMTarget Model=Gemini-2.5-flash2026.02564438
PROBELLMTarget Model=phi-42026.02534769
PROBELLMTarget Model=GPT-4o-mini2026.02475356
PROBELLMTarget Model=ministral-14b2026.02475364
PROBELLMTarget Model=Grok-4.1-fast2026.02455539
PROBELLMTarget Model=GPT-oss-20b2026.02455540