Mathematical Reasoning on Mathematical Reasoning Aggregate Average of 6 Benchmarks (Mean@8)
65.5Mean@8 ScoreDISA
Evaluation Results
| Method | Links | |
|---|---|---|
| DISABackbone=Qwen3-4B-Base2026.05 | 65.5 | |
| FlowRLBackbone=Qwen3-4B-Base2026.05 | 64.8 | |
| GSPOBackbone=Qwen3-4B-Base2026.05 | 62.7 | |
| GRPOBackbone=Qwen3-4B-Base2026.05 | 49.4 | |
| DISABackbone=Qwen2.5-7B2026.05 | 33.5 | |
| FlowRLBackbone=Qwen2.5-7B2026.05 | 33.2 | |
| GRPOBackbone=Qwen2.5-7B2026.05 | 32.8 | |
| VanillaBackbone=Qwen3-4B-Base2026.05 | 29.7 | |
| GSPOBackbone=Qwen2.5-7B2026.05 | 29.3 | |
| VanillaBackbone=Qwen2.5-7B2026.05 | 23.4 |