Mathematical Reasoning on MATH500 (Acc, NFE)
38.6Accuracy (MATH500)SWD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| SWDBackbone=Dream-7B, Scoring Metric=Entropy, Decoding Strategy=Stability-Weighted Decoding, lambda=52026.04 | 38.6 | 181.6 | |
| SWDBackbone=Dream-7B, Scoring Metric=Confidence, Decoding Strategy=Stability-Weighted Decoding, lambda=52026.04 | 35 | 151.6 | |
| SWDBackbone=Dream-7B, Scoring Metric=Margin, Decoding Strategy=Stability-Weighted Decoding, lambda=52026.04 | 33.8 | 135.4 | |
| SWDBackbone=LLaDA-8B, Scoring Metric=Confidence, Decoding Strategy=Stability-Weighted Decoding, lambda=52026.04 | 26.2 | 205.7 | |
| SWDBackbone=LLaDA-8B, Scoring Metric=Margin, Decoding Strategy=Stability-Weighted Decoding, lambda=52026.04 | 25.8 | 198.4 | |
| BaseBackbone=LLaDA-8B, Scoring Metric=Margin, Decoding Strategy=Vanilla EB Sampler2026.04 | 21.8 | 221 | |
| SWDBackbone=LLaDA-8B, Scoring Metric=Entropy, Decoding Strategy=Stability-Weighted Decoding, lambda=52026.04 | 20.4 | 224.3 | |
| BaseBackbone=LLaDA-8B, Scoring Metric=Confidence, Decoding Strategy=Vanilla EB Sampler2026.04 | 19.6 | 226.6 | |
| BaseBackbone=LLaDA-8B, Scoring Metric=Entropy, Decoding Strategy=Vanilla EB Sampler2026.04 | 19.6 | 230.7 | |
| BaseBackbone=Dream-7B, Scoring Metric=Margin, Decoding Strategy=Vanilla EB Sampler2026.04 | 11.4 | 354.9 | |
| BaseBackbone=Dream-7B, Scoring Metric=Confidence, Decoding Strategy=Vanilla EB Sampler2026.04 | 10.8 | 364.1 | |
| BaseBackbone=Dream-7B, Scoring Metric=Entropy, Decoding Strategy=Vanilla EB Sampler2026.04 | 8.4 | 382.4 |