Anomaly Reasoning on MMAD
91.2Average ScoreGLLS
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| GLLSBase Model=Qwen3-VL, Shot-setting=1-shot, Param=8B2026.07 | 91.2 | — | — | — | 96.5 | 87.7 | 89.3 | 86.9 | 95.5 | — | — | |
| Gemini-32026.07 | 89.3 | — | — | — | 85.7 | 86.1 | 88.9 | 89.9 | 95.7 | — | — | |
| GLLSBase Model=Qwen3-VL, Shot-setting=1-shot, Param=4B2026.07 | 87.7 | — | — | — | 96.4 | 84.3 | 83.2 | 81.9 | 92.5 | — | — | |
| GLLSBase Model=Qwen3-VL, Shot-setting=0-shot, Param=8B2026.07 | 87.1 | — | — | — | 88.6 | 81.9 | 86.2 | 85 | 93.9 | — | — | |
| Human (expert)Shot=1-shot2026.05 | 86.65 | — | — | — | 95.24 | 75 | 92.31 | 83.33 | 94.2 | 86.11 | 80.37 | |
| GLLSBase Model=Qwen2.5-VL, Shot-setting=1-shot, Param=7B2026.07 | 85.5 | — | — | — | 93.1 | 78.2 | 79.8 | 83.5 | 92.7 | — | — | |
| Qwen3-VLParam=8B2026.07 | 84.4 | — | — | — | 83.1 | 74.2 | 86.6 | 84.8 | 93.5 | — | — | |
| MiniCPM-V4Param=4B2026.07 | 83.1 | — | — | — | 79.4 | 68 | 86.2 | 87.2 | 94.5 | — | — | |
| AgentIADScale=3B, Shot=1-shot2026.05 | 82.88 | — | — | — | 69.49 | 72.73 | 80.94 | 85.27 | 87.84 | 93.27 | 90.59 | |
| MiniCPM-V4.5Param=9B2026.07 | 82.7 | — | — | — | 77.8 | 72.5 | 80.3 | 89.4 | 93.5 | — | — | |
| AD-Copilot-thinkingScale=7B, Shot=1-shot2026.05 | 82.29 | — | — | — | 73.95 | 74.29 | 76.4 | 84.92 | 86.93 | 91.86 | 87.78 | |
| Qwen3-VL-ThinkingParam=8B2026.07 | 82.1 | — | — | — | 81.7 | 69.7 | 81.7 | 83.5 | 93.7 | — | — | |
| GLLSBase Model=Qwen2.5-VL, Shot-setting=0-shot, Param=7B2026.07 | 82.1 | — | — | — | 84 | 73.3 | 77.4 | 83.6 | 92.4 | — | — | |
| GPT-5-mini2026.07 | 81.7 | — | — | — | 70.6 | 77.5 | 82.7 | 84 | 93.5 | — | — | |
| GLLSBase Model=Qwen3-VL, Shot-setting=0-shot, Param=4B2026.07 | 81.6 | — | — | — | 82.6 | 79.8 | 77.1 | 77.3 | 91.3 | — | — | |
| AnomalyClaw (GPT-5.5)Shot=1-shot2026.05 | 79.15 | — | — | — | 72.51 | 63.64 | 67.73 | 79.65 | 87.99 | 95.38 | 87.13 | |
| Qwen2.5-VLParam=7B2026.07 | 79 | — | — | — | 75.8 | 67.3 | 77.2 | 82.2 | 92.3 | — | — | |
| Gemini-2.52026.07 | 78.8 | — | — | — | 71.8 | 72.4 | 80.6 | 78.7 | 90.3 | — | — | |
| Human (ordinary)Shot=1-shot2026.05 | 78.69 | — | — | — | 86.9 | 66.25 | 85.58 | 71.25 | 81.52 | 89.58 | 69.72 | |
| Qwen2.5-VLScale=72B, Shot=1-shot2026.05 | 76.96 | — | — | — | 72.66 | 62.31 | 67.16 | 73.56 | 81.95 | 94.3 | 86.78 | |
| InternVL-3.5Param=8B2026.07 | 76.4 | — | — | — | 66.8 | 64.5 | 75.9 | 81.6 | 93.3 | — | — | |
| GPT-5.5Shot=1-shot2026.05 | 76.2 | — | — | — | 69.95 | 64.91 | 62.34 | 75.95 | 80.1 | 95.23 | 82.11 | |
| GPT-5-nano2026.07 | 75.7 | — | — | — | 73.5 | 63.8 | 73.9 | 75.7 | 91.8 | — | — | |
| Seed1.5-VLShot=1-shot2026.05 | 75.21 | — | — | — | 65.3 | 64.32 | 63.35 | 73.8 | 84.38 | 91.65 | 83.67 | |
| Qwen3-VLScale=8B, Shot=1-shot2026.05 | 73.91 | — | — | — | 68.71 | 63.27 | 61.01 | 68.18 | 79.99 | 92.27 | 83.96 | |
| InternVL-3.5Param=4B2026.07 | 73.9 | — | — | — | 74.4 | 57.9 | 67.7 | 78.2 | 91.5 | — | — | |
| LLaVA-OVSeek-Setting=true2025.08 | 73.54 | 82.77 | 70.31 | 67.55 | — | — | — | — | — | — | — | |
| Qwen2.5-VLParam=3B2026.07 | 73.5 | — | — | — | 68.3 | 58.5 | 72.5 | 76.8 | 91.5 | — | — | |
| Gemini-1.5-proShot=1-shot2026.05 | 73.09 | — | — | — | 68.63 | 60.12 | 58.56 | 70.38 | 82.46 | 89.2 | 82.25 | |
| Qwen3-VLParam=4B2026.07 | 72.4 | — | — | — | 70.5 | 64.5 | 79 | 57.5 | 90.5 | — | — | |
| VLM R1Seek-Setting=true2025.08 | 72.26 | 77.65 | 71.03 | 68.11 | — | — | — | — | — | — | — | |
| Anomaly R1training-based=true2025.08 | 71.93 | 81.99 | 70.55 | 63.27 | — | — | — | — | — | — | — | |
| InternVL2Scale=76B, Shot=1-shot2026.05 | 70.75 | — | — | — | 68.25 | 54.22 | 56.66 | 66.3 | 80.47 | 86.4 | 82.92 | |
| Anomaly OVtraining-based=true2025.08 | 69.8 | 73.99 | 65.39 | 70.02 | — | — | — | — | — | — | — | |
| EIADScale=8B, Shot=1-shot2026.05 | 69.4 | — | — | — | 60.5 | 50.7 | 55.6 | 67.8 | 76.9 | 91.8 | 82.7 | |
| VLM R1Seek-Setting=false2025.08 | 64.69 | 72.01 | 58.47 | 63.61 | — | — | — | — | — | — | — | |
| LLaVA-OVSeek-Setting=false2025.08 | 64.13 | 70.13 | 63.21 | 59.05 | — | — | — | — | — | — | — | |
| AnomalyGPTScale=7B, Shot=1-shot2026.05 | 36.52 | — | — | — | 65.57 | 27.49 | 27.97 | 36.86 | 32.11 | 29.84 | 35.82 |