Tag Prediction on HPR-Benchtag
73.29Average ScoreQwen3.6-Plus
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| Qwen3.6-PlusModel Category=Thinking Models2026.05 | 73.29 | 79.07 | 59.65 | 69.91 | 66.94 | 79.93 | |
| UserGPTSFT# Params=Dense (8B)2026.05 | 73.25 | 77.76 | 60.88 | 69.99 | 82.12 | 75.48 | |
| UserGPT# Params=Dense (8B)2026.05 | 73.06 | 78.79 | 59.62 | 69.45 | 81.64 | 75.46 | |
| GLM-5Model Category=Thinking Models, # Params=MoE (744B/A40B)2026.05 | 71.86 | 76.91 | 53.45 | 70.98 | 70.37 | 78.38 | |
| Kimi-K2.5Model Category=Thinking Models, # Params=MoE (1T/A32B)2026.05 | 70.64 | 74.67 | 53.8 | 68.66 | 65.9 | 78.34 | |
| DeepSeek-V3.2Model Category=Thinking Models, # Params=MoE (671B/37B)2026.05 | 66.84 | 68.42 | 49.81 | 67.53 | 66.93 | 73.14 | |
| DeepSeek-R1-0528Model Category=Thinking Models, # Params=MoE (671B/37B)2026.05 | 66.13 | 64.12 | 47.52 | 66.64 | 67.86 | 74.24 | |
| Qwen3-235B-A22B-Thinking-2507Model Category=Thinking Models, # Params=MoE (235B/A22B)2026.05 | 64.34 | 64.11 | 41.35 | 64.38 | 62.61 | 74.78 | |
| DNNModel Category=Discriminative Models2026.05 | 58.46 | 71.56 | 40 | 65.25 | 56.16 | — | |
| Qwen3-235B-A22B-Instruct-2507Model Category=Non-Thinking Models, # Params=MoE (235B/A22B)2026.05 | 52.44 | 59.44 | 41.4 | 51.94 | 55.84 | 53.41 | |
| Qwen3-8B ThinkingModel Category=Thinking Models, # Params=Dense (8B)2026.05 | 50.35 | 62.24 | 38.31 | 62.44 | 47.7 | 45.99 | |
| Qwen3-8B Non-ThinkingModel Category=Non-Thinking Models, # Params=Dense (8B)2026.05 | 46.12 | 58.07 | 32.81 | 60.91 | 42.82 | 41.3 |