Mathematical Reasoning on AMC23 (Accuracy (%))
72.5Accuracy (%)MCTS (Vanilla)
Evaluation Results
| Method | Links | |
|---|---|---|
| MCTS (Vanilla)Base Model=Qwen2026.04 | 72.5 | |
| MCTS + PEBase Model=Qwen2026.04 | 69.2 | |
| Beam SearchBase Model=Qwen2026.04 | 67.5 | |
| Adaptive Parallel Monte Carlo Tree Search (Full Suite)Base Model=Qwen2026.04 | 65 | |
| MCTS (Vanilla)Base Model=Llama2026.04 | 57.5 | |
| Adaptive Parallel Monte Carlo Tree Search (Full Suite)Base Model=Llama2026.04 | 55 | |
| Beam SearchBase Model=Llama2026.04 | 52.5 | |
| MCTS + PEBase Model=Llama2026.04 | 47.5 |