Mathematical Reasoning on AIMO 2025 (Pass@32)
49.08Pass@32EAPO
Evaluation results
| Method | Links | ||
|---|---|---|---|
| EAPOPolicy Backbone=Qwen3-8B2025.09 | 49.08 | 0.0743 | |
| Self-Exploratory RLPolicy Backbone=Qwen3-8B2025.09 | 42.59 | 0.0811 | |
| Expert-Assisted WorkflowPolicy Backbone=Qwen3-8B2025.09 | 30.01 | 0.2224 | |
| Base ModelPolicy Backbone=Qwen3-8B2025.09 | 23.57 | 0.1281 |