Mathematical Reasoning on AMC 23 (Average@16)
66.56Average@16LENS_GRPO
Evaluation Results
| Method | Links | |
|---|---|---|
| LENS_GRPOBase Model=Qwen3-8B-Base2026.01 | 66.56 | |
| LENS_GRPOBase Model=Qwen3-4B-Base2026.01 | 60.94 | |
| Qwen3-8B-BaseBase Model=Qwen3-8B-Base2026.01 | 38.59 | |
| LENS_GRPOBase Model=Llama3.2-3B-Instruct2026.01 | 29.84 | |
| Qwen3-4B-BaseBase Model=Qwen3-4B-Base2026.01 | 26.72 | |
| Llama3.2-3B-InstructBase Model=Llama3.2-3B-Instruct2026.01 | 10.16 |