Target Behavior Attack on MMLU phy
99.3ASRAiTM
Evaluation results
| Method | Links | |
|---|---|---|
| AiTMFramework=Camel, Communication Structure=Chain2025.02 | 99.3 | |
| AiTMFramework=Camel, Communication Structure=Complete2025.02 | 97.1 | |
| AiTMFramework=AutoGen, Communication Structure=Chain2025.02 | 90.1 | |
| AiTMFramework=AutoGen, Communication Structure=Complete2025.02 | 89.4 | |
| AiTMFramework=AutoGen, Communication Structure=Chain2025.02 | 87.6 | |
| AiTMFramework=Camel, Communication Structure=Tree2025.02 | 85.7 | |
| AiTMFramework=AutoGen, Communication Structure=Tree2025.02 | 79.5 | |
| AiTMFramework=Camel, Communication Structure=Random2025.02 | 79.4 | |
| AiTMFramework=Camel, Communication Structure=Chain2025.02 | 77.4 | |
| AiTMFramework=Camel, Communication Structure=Complete2025.02 | 72.6 | |
| AiTMFramework=AutoGen, Communication Structure=Random2025.02 | 70.8 | |
| AiTMFramework=Camel, Communication Structure=Random2025.02 | 61.2 | |
| AiTMFramework=Camel, Communication Structure=Tree2025.02 | 52.3 | |
| AiTMFramework=AutoGen, Communication Structure=Complete2025.02 | 50.9 | |
| AiTMFramework=AutoGen, Communication Structure=Tree2025.02 | 45.4 | |
| AiTMFramework=AutoGen, Communication Structure=Random2025.02 | 40.8 |