Safety Detection on MAS LLaMA-3.1-8B (Lookahead)
92.31AccuracyAgentLens
Evaluation Results
| Method | Links | |
|---|---|---|
| AgentLensMethod Category=Interpretability-based2026.06 | 92.31 | |
| Claude Sonnet-4.6Method Category=LLM-as-a-Judge2026.06 | 84.62 | |
| GPT-4oMethod Category=LLM-as-a-Judge2026.06 | 69.23 | |
| Claude Opus-4.6Method Category=LLM-as-a-Judge2026.06 | 69.23 | |
| GPT-5.2Method Category=LLM-as-a-Judge2026.06 | 61.54 | |
| Deepseek-V4-ProMethod Category=LLM-as-a-Judge2026.06 | 15.38 | |
| Gemini-3-ProMethod Category=LLM-as-a-Judge2026.06 | 7.69 | |
| Deepseek-V4-FlashMethod Category=LLM-as-a-Judge2026.06 | 7.69 |