Mathematical Reasoning on MATH (Accuracy, TPS, and Speedup)
38.24AccuracyConfidence
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| ConfidenceModel=Dream-v0-Instruct-7B2026.02 | 38.24 | 37.92 | 2.14 | |
| DAWNModel=Dream-v0-Instruct-7B2026.02 | 38.22 | 44.18 | 2.49 | |
| LocalLeapModel=Dream-v0-Instruct-7B2026.02 | 38.1 | 43.66 | 2.46 | |
| OriginalModel=Dream-v0-Instruct-7B2026.02 | 37.8 | 17.73 | 1 | |
| OriginalModel=Dream-v0-Base-7B2026.02 | 34.7 | 18.16 | 1 | |
| KLASSModel=Dream-v0-Instruct-7B2026.02 | 34.66 | 18.74 | 1.06 | |
| LocalLeapModel=Dream-v0-Base-7B2026.02 | 34.48 | 51.32 | 2.83 | |
| ConfidenceModel=Dream-v0-Base-7B2026.02 | 34.38 | 44.66 | 2.46 | |
| DAWNModel=Dream-v0-Base-7B2026.02 | 34 | 51.45 | 2.83 | |
| ConfidenceModel=LLaDA-1.52026.02 | 33.38 | 32.87 | 2.58 | |
| OriginalModel=LLaDA-1.52026.02 | 33.36 | 12.75 | 1 | |
| OriginalModel=LLaDA-8B-Instruct2026.02 | 33.2 | 14.1 | 1 | |
| ConfidenceModel=LLaDA-8B-Instruct2026.02 | 32.98 | 36.88 | 2.62 | |
| KLASSModel=Dream-v0-Base-7B2026.02 | 32.8 | 23.16 | 1.27 | |
| LocalLeapModel=LLaDA-1.52026.02 | 32.44 | 42.36 | 3.32 | |
| LocalLeapModel=LLaDA-8B-Instruct2026.02 | 32.42 | 47.12 | 3.34 | |
| DAWNModel=LLaDA-8B-Instruct2026.02 | 32.36 | 48.16 | 3.42 | |
| KLASSModel=LLaDA-8B-Instruct2026.02 | 32.32 | 25.27 | 1.79 | |
| KLASSModel=LLaDA-1.52026.02 | 31.8 | 23.29 | 1.83 | |
| DAWNModel=LLaDA-1.52026.02 | 31.8 | 42.84 | 3.36 |