Mathematical Reasoning on AIME 25 (Maj, Debate, Delta)
28.7Maj ScoreSDRL-freq
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SDRL-freqBackbone=Qwen3-4B-Base, Debate system agents (N)=5, Debate rounds=12026.01 | 28.7 | 30 | 1.3 | |
| DAPOBackbone=Qwen3-4B-Base, Debate system agents (N)=5, Debate rounds=12026.01 | 24.7 | 24 | -0.7 | |
| SDRL-randBackbone=Qwen3-4B-Base, Debate system agents (N)=5, Debate rounds=12026.01 | 22.7 | 24 | 1.3 | |
| SDRL-randBackbone=Qwen2.5-3B, Debate system agents (N)=5, Debate rounds=12026.01 | 5.6 | 6.1 | 0.5 | |
| DAPOBackbone=Qwen2.5-3B, Debate system agents (N)=5, Debate rounds=12026.01 | 4 | 3.3 | -0.7 | |
| SDRL-freqBackbone=Qwen2.5-3B, Debate system agents (N)=5, Debate rounds=12026.01 | 3.3 | 4.7 | 1.4 |