Mathematical Reasoning on HMMT 24 (It0, It1, It2, It3)
26.8Iteration 0 ScoreBase
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| BaseBackbone=Qwen3-30B-Instruct2026.01 | 26.8 | — | — | — | |
| Self-RefBackbone=Qwen3-30B-Instruct2026.01 | 26.8 | 39.5 | 41.6 | 43.7 | |
| PaCoReBackbone=Qwen3-30B-Instruct2026.01 | 26.8 | 52.8 | 53.6 | 53.7 | |
| RSEBackbone=Qwen3-30B-Instruct2026.01 | 26.8 | 50.3 | 55.4 | 54.4 | |
| MV@128Backbone=Qwen3-30B-Instruct2026.01 | 23.7 | — | — | — | |
| MV@128Backbone=Qwen3-4B-Instruct2026.01 | 21.8 | — | — | — | |
| BaseBackbone=Qwen3-4B-Instruct2026.01 | 21.7 | — | — | — | |
| Self-RefBackbone=Qwen3-4B-Instruct2026.01 | 21.7 | 26.8 | 28.6 | 29.4 | |
| PaCoReBackbone=Qwen3-4B-Instruct2026.01 | 21.7 | 35 | 38.2 | 40 | |
| RSEBackbone=Qwen3-4B-Instruct2026.01 | 21.7 | 37.8 | 40.4 | 41.5 |