Knowledge Reasoning on GPQA
67.3AccuracyLLaDA-2.1-flash
Evaluation Results
| Method | Links | |
|---|---|---|
| LLaDA-2.1-flashParams=100B2026.04 | 67.3 | |
| Qwen3Params=32B2026.04 | 65 | |
| LLaDA-2.0-flashParams=100B2026.04 | 62.3 | |
| I-DLMParams=32B, Decoding=ISD (N=4, sampling)2026.04 | 58.7 | |
| ERSSModel=Phi-4-Reasoning2026.06 | 56.7 | |
| EASDTM=32B, DM=7B2025.12 | 55.55 | |
| LRSModel=Phi-4-Reasoning2026.06 | 55.47 | |
| Qwen3Params=8B2026.04 | 55.4 | |
| ERSSModel=R1-Distilled-LLaMA-8B2026.06 | 55.13 | |
| I-DLMParams=8B, Decoding=ISD (N=4, sampling)2026.04 | 54.9 | |
| ALRRModel=Phi-4-Reasoning2026.06 | 54.69 | |
| LRSModel=R1-Distilled-LLaMA-8B2026.06 | 54.68 | |
| BaselineModel=Phi-4-Reasoning2026.06 | 53.79 | |
| LLaDA-2.1-miniParams=16B2026.04 | 53.3 | |
| ALRR+ERSSModel=R1-Distilled-LLaMA-8B2026.06 | 53.12 | |
| ALRRModel=R1-Distilled-LLaMA-8B2026.06 | 52.23 | |
| ALRR+LRSModel=R1-Distilled-LLaMA-8B2026.06 | 52.12 | |
| EASDTM=72B, DM=7B2025.12 | 52.02 | |
| ALRRModel=Qwen3-4B2026.06 | 51.55 | |
| ERSSModel=Qwen3-4B2026.06 | 51.33 | |
| ALRR+ERSSModel=Qwen3-4B2026.06 | 50.54 | |
| RSDTM=32B, DM=7B, PRM=1.5B2025.12 | 50.5 | |
| SDTM=72B, DM=7B2025.12 | 50.5 | |
| BaselineModel=R1-Distilled-LLaMA-8B2026.06 | 50.45 | |
| BaselineModel=Qwen3-4B2026.06 | 50.22 | |
| LRSModel=Qwen3-4B2026.06 | 50.22 | |
| ERSSModel=Phi-42026.06 | 49.55 | |
| SDTM=32B, DM=7B2025.12 | 49.49 | |
| ALRR+LRSModel=Qwen3-4B2026.06 | 49.41 | |
| ALRRModel=Phi-42026.06 | 49.11 | |
| LRSModel=Phi-42026.06 | 48.55 | |
| BaselineModel=Phi-42026.06 | 48.21 | |
| ERSS+LRSModel=R1-Distilled-LLaMA-8B2026.06 | 48.21 | |
| RSDTM=72B, DM=7B, PRM=1.5B2025.12 | 47.48 | |
| Single ModelTM=72B2025.12 | 46.96 | |
| ERSS+LRSModel=Qwen3-4B2026.06 | 46.88 | |
| Single ModelTM=32B2025.12 | 45.96 | |
| Beam Search (N=16)DM=7B2025.12 | 42.42 | |
| Majority Voting (N=16)DM=7B2025.12 | 41.41 | |
| ERSSModel=Gemma3-12B-it2026.06 | 38.62 | |
| Single ModelDM=7B2025.12 | 36.36 | |
| ALRRModel=Gemma3-12B-it2026.06 | 36.05 | |
| LRSModel=Gemma3-12B-it2026.06 | 36.05 | |
| BaselineModel=Gemma3-12B-it2026.06 | 33.26 | |
| LRSModel=Gemma3-4B-it2026.06 | 27.23 | |
| ERSSModel=Gemma3-4B-it2026.06 | 27.12 | |
| BaselineModel=Gemma3-4B-it2026.06 | 26.67 | |
| ALRRModel=Gemma3-4B-it2026.06 | 26.45 |