Resource Extraction on DeepLabCut (test)
100Primary Fields ScoreStructSense
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| StructSenseModel=GPT-4o-mini2025.07 | 100 | 75 | 94 | 88.9 | |
| StructSenseModel=Gemini 3.1 Flash Lite2025.07 | 85 | 100 | 96 | 93.2 | |
| StructSenseModel=Qwen2025.07 | 75 | 100 | 98 | 89.4 |