Exfiltration on ACIARENA Math
62.22UALLM Debate
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| LLM DebateBase LLM=GPT-4o-mini2026.04 | 62.22 | 57.27 | |
| Self ConsistencyBase LLM=GPT-4o-mini2026.04 | 59.49 | 43.59 | |
| AutoGenBase LLM=GPT-4o-mini2026.04 | 58.12 | 48.38 | |
| AgentVerseBase LLM=GPT-4o-mini2026.04 | 55.21 | 40.51 | |
| CAMELBase LLM=GPT-4o-mini2026.04 | 33.5 | 22.56 |