Recognition on DeepTumorVQA refined subset 2025b
82Colon Lesion Existence3DMedAgent (GPT-5)
Evaluation Results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| 3DMedAgent (GPT-5)Base Model=GPT-5, Evaluation Protocol=zero-shot2026.02 | 82 | 82 | 63 | 85 | 73 | 72 | 76 | |
| 3DMedAgent (Qwen3-VL)Base Model=Qwen3-VL (30B), Evaluation Protocol=zero-shot2026.02 | 80 | 78 | 55 | 82 | 75 | 68 | 73 | |
| M3DEvaluation Protocol=zero-shot2026.02 | 60 | 57 | 44 | 52 | 60 | 49 | 53 | |
| HuatuoGPTParameters=34B, Evaluation Protocol=zero-shot2026.02 | 57 | 57 | 56 | 53 | 43 | 60 | 54 | |
| RandType=Random baseline2026.02 | 50 | 50 | 50 | 50 | 50 | 50 | 50 | |
| MedGemmaParameters=27B, Evaluation Protocol=zero-shot2026.02 | 50 | 50 | 49 | 57 | 50 | 49 | 51 | |
| RadFMEvaluation Protocol=zero-shot2026.02 | 48 | 48 | 41 | 55 | 43 | 49 | 47 | |
| GPT-5Evaluation Protocol=zero-shot2026.02 | 47 | 57 | 46 | 58 | 38 | 53 | 49 | |
| Qwen3-VLParameters=30B, Evaluation Protocol=zero-shot2026.02 | 47 | 48 | 47 | 52 | 47 | 53 | 51 |