Autonomous LLM Agent Verification on OPERA
11.9Mean Td (s)VFA
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| VFASystem Variant=VFA (Ours)2025.12 | 11.9 | 9.2 | 85 | 9 | 72 | |
| Log-Monitoring (Heuristic)System Variant=Log-Monitoring (Heuristic)2025.12 | 21.8 | 11.3 | 73 | 12 | 69 | |
| No-Verifier BaselineSystem Variant=No-Verifier Baseline2025.12 | 35.4 | 18.9 | 62 | 15 | 58 |