merck-p2
Intersection F1
0.890
Fiducial F1
1.000
Label accuracy
1.000
Leader accuracy
0.077
Trace IoU
0.693
End-to-end F1
0.538
Seconds / PDF
7.9s
Metric comparison (normalized)
Each metric scaled to [0,1] so you can see correlation and co-movement
Evolution across milestones
Same pdf evaluated at each milestone
| Milestone | Intersection F1 | Fiducial F1 | Label | Leader | Trace IoU | e2e F1 | Seconds / PDF |
|---|---|---|---|---|---|---|---|
| m01-v2-weights | 0.890 | 1.000 | 1.000 | 0.077 | 0.693 | 0.538 | 7.9s |
| m02-v3-weights | 0.520 | 1.000 | 1.000 | 0.077 | 0.499 | 0.192 | 5.1s |
| m03-v4-weights | 0.994 | 1.000 | 1.000 | 0.577 | 0.879 | 0.808 | 10.1s |
| m04-classifier | 0.994 | 1.000 | 1.000 | 1.000 | 0.916 | 0.923 | 9.2s |
| m05-llm-initial | 0.994 | 1.000 | 1.000 | 1.000 | 0.916 | 0.923 | 13.3s |
| m06-grid-bbox | 0.994 | 1.000 | 1.000 | 1.000 | 0.916 | 0.923 | 10.8s |
| m07-full-detector | 0.994 | 1.000 | 1.000 | 1.000 | 0.916 | 0.923 | 10.7s |
| m08-spatial-index | 0.548 | 0.762 | 1.000 | 1.000 | 0.925 | 0.714 | 7.8s |
| m09-current-logic | 0.994 | 1.000 | 1.000 | 1.000 | 0.916 | 0.923 | 9.7s |
| m10-v5-weights | 0.978 | 1.000 | 1.000 | 1.000 | 0.979 | 0.962 | 10.8s |
| m11-latest-pipeline | 0.978 | 1.000 | 1.000 | 1.000 | 0.979 | 0.962 | 9.5s |
| auto-fbce4876 | 0.944 | 0.939 | 1.000 | 1.000 | 0.976 | 0.898 | 6.2s |
| auto-b62c0378 | 0.944 | 0.939 | 1.000 | 1.000 | 0.976 | 0.898 | 6.5s |