Automated Probing on MBPP
0.38Error Rate (%)PAIR
Evaluation results
| Method | Links | ||
|---|---|---|---|
| PAIRGenerator Model=GPT-5.2, Target Model=Llama-3.1-8b-ins.2026.02 | 0.38 | 94.85 | |
| AutoDetectGenerator Model=GPT-5.2, Target Model=Llama-3.1-8b-ins.2026.02 | 0.46 | — | |
| PROBELLMGenerator Model=GPT-5.2, Target Model=Llama-3.1-8b-ins.2026.02 | 0.9 | — |