Grade School Math Word Problems on GSM8K (256 tokens context)
81.2Pass@1d-TreeRPO
Evaluation Results
| Method | Links | |
|---|---|---|
| d-TreeRPOBase Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 81.2 | |
| GDPOBase Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 81.1 | |
| wd1Base Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 80.8 | |
| SAPOBase Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 80.6 | |
| TraceRLBase Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 80.3 | |
| VRPO (LLaDA-1.5)Base Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 80.1 | |
| Diffu-GRPOBase Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 79.8 | |
| LLaDA-8B-InstructBase Model=LLaDA-8B-Instruct, Generation Length=2562025.12 | 76.7 | |
| wd1Base Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 76.2 | |
| d-TreeRPOBase Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 75.5 | |
| TraceRLBase Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 75.4 | |
| SAPOBase Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 75.2 | |
| GDPOBase Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 74.8 | |
| Diffu-GRPOBase Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 74.2 | |
| LLaDA-MoE-7BA1B-InstructBase Model=LLaDA-MoE-7BA1B-Instruct, Generation Length=2562025.12 | 71.8 |