Multi-agent debate on Mixed regime
72.4AccuracyFlat
Evaluation Results
| Method | Links | |
|---|---|---|
| FlatModel=gpt-oss-120b2026.06 | 72.4 | |
| MADModel=gpt-oss-120b2026.06 | 72.2 | |
| Transformational leadership controllerModel=gpt-oss-120b2026.06 | 71.8 | |
| Situational leadership controllerModel=gpt-oss-120b2026.06 | 71.6 | |
| R0-voteModel=gpt-oss-120b2026.06 | 71.3 | |
| MADModel=llama-4-scout2026.06 | 71.3 | |
| FlatModel=llama-4-scout2026.06 | 71.1 | |
| Situational leadership controllerModel=llama-4-scout2026.06 | 70.7 | |
| Transformational leadership controllerModel=llama-4-scout2026.06 | 69.6 | |
| Situational leadership controllerModel=gemma-4-31B-it2026.06 | 69.6 | |
| R0-voteModel=llama-4-scout2026.06 | 69.3 | |
| Transformational leadership controllerModel=gemma-4-31B-it2026.06 | 69.1 | |
| MADModel=gemma-4-31B-it2026.06 | 68 | |
| R0-voteModel=gemma-4-31B-it2026.06 | 67.8 | |
| FlatModel=gemma-4-31B-it2026.06 | 67.1 |