Patient-level agreement on Open-Weight (OW) cohort (full)
0.953Quadratic Weighted Kappa (kw)Gemma vs Qwen
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemma vs QwenContext=OW inter-assessor, N=4,7692026.07 | 0.953 | |
| Assessors (A1, A2) vs consolidated outputContext=OW internal pipeline agreement, N=4,770–4,7912026.07 | 0.911 | |
| Cooper GT vs Young GTContext=GT inter-agreement, N=4,8862026.07 | 0.819 | |
| OW MOSAICComparison Reference=Cooper GT, N=4,8862026.07 | 0.597 | |
| Young GT vs DiSSCo GTContext=GT inter-agreement, N=4,8862026.07 | 0.585 | |
| OW MOSAICComparison Reference=Young GT, N=4,8862026.07 | 0.534 | |
| Cooper GT vs DiSSCo GTContext=GT inter-agreement, N=4,8862026.07 | 0.431 | |
| OW MOSAICComparison Reference=MOSAIC Frozen, N=4,8862026.07 | 0.428 | |
| OW MOSAICComparison Reference=DiSSCo GT, N=4,8862026.07 | 0.32 |