Vision-Language Reasoning on CheXthought 300 1.0 (test)
4.67Comprehensiveness of FindingsCheXthought Data
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| CheXthought Data2026.04 | 4.67 | 4.84 | 4.88 | 4.88 | 4.81 | |
| GPT 5.22026.04 | 4.25 | 4.37 | 4.19 | 3.79 | 4.14 | |
| MedGemma 1.52026.04 | 4.17 | 3.99 | 4.12 | 4.34 | 4.13 | |
| Claude Opus 4.52026.04 | 4.03 | 4.26 | 3.83 | 3.5 | 3.89 |