General Reasoning on CSQA
81.3AccuracyVLP
Evaluation results
| Method | Links | ||
|---|---|---|---|
| VLPBackbone=Qwen3-14B, Sampling Strategy=mean@82026.03 | 81.3 | 220 | |
| CODABackbone=Qwen3-14B, Sampling Strategy=mean@82026.03 | 81.3 | 206 | |
| ASRRBackbone=Qwen3-14B, Sampling Strategy=mean@82026.03 | 81.2 | 163 | |
| GRPOBackbone=Qwen3-14B, Sampling Strategy=mean@82026.03 | 80.7 | 496 | |
| CODABackbone=Qwen3-8B, Sampling Strategy=mean@82026.03 | 79.9 | 139 | |
| GRPOBackbone=Qwen3-8B, Sampling Strategy=mean@82026.03 | 79.8 | 571 | |
| ASRRBackbone=Qwen3-8B, Sampling Strategy=mean@82026.03 | 79.7 | 203 | |
| VLPBackbone=Qwen3-8B, Sampling Strategy=mean@82026.03 | 79.3 | 267 | |
| CODABackbone=Qwen3-4B, Sampling Strategy=mean@82026.03 | 78.9 | 42 | |
| Qwen3-14B-BaseBackbone=Qwen3-14B, Sampling Strategy=mean@82026.03 | 78.8 | 214 | |
| VLPBackbone=Qwen3-4B, Sampling Strategy=mean@82026.03 | 78.3 | 118 | |
| ASRRBackbone=Qwen3-4B, Sampling Strategy=mean@82026.03 | 78 | 185 | |
| GRPOBackbone=Qwen3-4B, Sampling Strategy=mean@82026.03 | 77.7 | 492 | |
| Qwen3-8B-BaseBackbone=Qwen3-8B, Sampling Strategy=mean@82026.03 | 74.6 | 252 | |
| Qwen3-4B-BaseBackbone=Qwen3-4B, Sampling Strategy=mean@82026.03 | 58.4 | 2,193 |