General Reasoning on GPQA (TPF, ACC)
314TPFLLaDA
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LLaDAModel Type=Masked Diffusion2026.02 | 314 | 17.2 | |
| LLaDA-1.5Model Type=Masked Diffusion2026.02 | 301 | 24.2 | |
| DiRL-8B-InstructModel Type=Block Diffusion, Block Size (B)=42026.02 | 227 | 44.4 | |
| LLaDA-MoEModel Type=Masked Diffusion2026.02 | 218 | 27.8 | |
| Fast-dLLM-v2Model Type=Block Diffusion, Block Size (B)=322026.02 | 209 | 28.3 | |
| TDAR-8B-thinking + BACDModel Type=Block Diffusion, Block Size (B)=16, Decoding=BACD2026.02 | 149 | 46 | |
| TraDo-8B-InstructModel Type=Block Diffusion, Block Size (B)=42026.02 | 143 | 27.3 | |
| TDAR-8B-thinking + BACD + TCCFModel Type=Block Diffusion, Block Size (B)=16, Decoding=BACD + TCCF2026.02 | 139 | 50 | |
| SDAR-8B-ChatModel Type=Block Diffusion, Block Size (B)=42026.02 | 132 | 10.6 | |
| TDAR-8B-thinkingModel Type=Block Diffusion, Block Size (B)=162026.02 | 128 | 46.5 | |
| TraDo-8B-Thinking + BACDModel Type=Block Diffusion, Block Size (B)=4, Decoding=BACD2026.02 | 118 | 49.5 | |
| TraDo-8B-ThinkingModel Type=Block Diffusion, Block Size (B)=42026.02 | 116 | 46 | |
| TraDo-8B-Thinking + BACD + TCCFModel Type=Block Diffusion, Block Size (B)=4, Decoding=BACD + TCCF2026.02 | 114 | 49.5 | |
| Qwen3-8B-Thinking†Model Type=Autoregressive2026.02 | 100 | 49 |