Mathematical Reasoning on MATH 500 (Accuracy, Steps, TPF)
40.6AccuracyKLASS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| KLASSBackbone=Dream2026.07 | 40.6 | 148.85 | 1.72 | |
| KLASS + TACGBackbone=Dream2026.07 | 40.6 | 145.79 | 1.76 | |
| Conf. + TACGBackbone=Dream2026.07 | 40 | 80.33 | 3.19 | |
| Conf.Backbone=Dream2026.07 | 39 | 94.6 | 2.71 | |
| d3LLM†Backbone=Dream2026.07 | 38.2 | — | 3.92 | |
| KLASS + TACGBackbone=LLaDA2026.07 | 33.8 | 133.33 | 2.43 | |
| Conf. + TACGBackbone=LLaDA2026.07 | 32.2 | 86.59 | 3.78 | |
| Conf.Backbone=LLaDA2026.07 | 30.6 | 96.48 | 3.37 | |
| KLASSBackbone=LLaDA2026.07 | 30.6 | 127.7 | 2.52 | |
| d3LLM†Backbone=LLaDA2026.07 | 30.4 | — | 5.74 |