Mathematical Reasoning on AIME 2025 (mean@30, pass@30)
26.8Mean Score @30FE-SFT & FC-RL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| FE-SFT & FC-RLBackbone=Youtu-LLM-2B-Base, Training Stage=RL, Framework=WM-AMT2026.06 | 26.8 | 53.3 | |
| FE-SFT & RL*Backbone=Youtu-LLM-2B-Base, Training Stage=RL2026.06 | 26.4 | 46.7 | |
| FE-SFT & RLBackbone=Youtu-LLM-2B-Base, Training Stage=RL, Framework=WM-AMT2026.06 | 25.6 | 53.3 | |
| IWM & RLBackbone=Youtu-LLM-2B-Base, Training Stage=RL2026.06 | 24.9 | 50 | |
| State-SFT & RL‡Backbone=Youtu-LLM-2B-Base, Training Stage=RL, Framework=WM-AMT2026.06 | 24.8 | 56.7 | |
| SFT & RL*Backbone=Youtu-LLM-2B-Base, Training Stage=RL2026.06 | 24.3 | 46.7 | |
| SFT & RLBackbone=Youtu-LLM-2B-Base, Training Stage=RL, Framework=WM-AMT2026.06 | 23.4 | 53.3 | |
| State-SFT & RL†Backbone=Youtu-LLM-2B-Base, Training Stage=RL2026.06 | 22.1 | 50 |