Systemic Risk Detection on EIA
94Action Grounding (Grd)LLaMA-Guard 3
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| LLaMA-Guard 3Framework Category=Guardrail-based2025.02 | 94 | 90 | 100 | |
| AgentMonitorFramework Category=Guardrail-based, Backbone=GPT-4o2025.02 | 58 | 40 | 100 | |
| GPT-4oFramework Category=Model-based, Shots=0-Shot2025.02 | 42 | 16 | 66.7 | |
| Claude-3.5Framework Category=Model-based, Shots=0-Shot2025.02 | 40 | 28 | 56.7 | |
| AGrailFramework Category=Guardrail-based, Backbone=GPT-4o2025.02 | 8 | 26 | 76.7 | |
| AGrailFramework Category=Guardrail-based, Backbone=Claude-3.52025.02 | 6 | 28 | 86.7 |