List your company

Automated Probing on HellaSwag

35Error RatePAIR
33.444.25565.8Feb 13, 2026

Evaluation results

MethodLinks
PAIRGenerator Model=GPT-5.2, Target Model=Llama-3.1-8b-ins.2026.023599.06
AutoDetectGenerator Model=GPT-5.2, Target Model=Llama-3.1-8b-ins.2026.0245—
PROBELLMGenerator Model=GPT-5.2, Target Model=Llama-3.1-8b-ins.2026.0275—