newman-p1
Intersection F1
0.974
Fiducial F1
1.000
Label accuracy
1.000
Leader accuracy
0.927
Trace IoU
0.898
End-to-end F1
0.857
Seconds / PDF
11.0s
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 | 0.920 | 0.808 | 1.000 | 0.488 | 0.845 | 0.654 | 27.4s |
| m02-v3-weights | 0.945 | 0.808 | 1.000 | 0.610 | 0.873 | 0.673 | 30.0s |
| m03-v4-weights | 0.970 | 0.808 | 1.000 | 0.268 | 0.903 | 0.731 | 35.5s |
| m04-classifier | 0.955 | 0.808 | 1.000 | 0.927 | 0.915 | 0.712 | 30.2s |
| m05-llm-initial | 0.960 | 1.000 | 1.000 | 0.927 | 0.915 | 0.881 | 31.0s |
| m06-grid-bbox | 0.960 | 1.000 | 1.000 | 0.927 | 0.915 | 0.881 | 30.7s |
| m07-full-detector | 0.960 | 1.000 | 1.000 | 0.927 | 0.915 | 0.881 | 29.5s |
| m08-spatial-index | 0.960 | 1.000 | 1.000 | 0.927 | 0.915 | 0.881 | 18.2s |
| m09-current-logic | 0.960 | 1.000 | 1.000 | 0.927 | 0.915 | 0.881 | 18.3s |
| m10-v5-weights | 0.954 | 1.000 | 1.000 | 0.927 | 0.903 | 0.857 | 19.5s |
| m11-latest-pipeline | 0.974 | 1.000 | 1.000 | 0.927 | 0.898 | 0.857 | 17.7s |
| auto-fbce4876 | 0.974 | 1.000 | 1.000 | 0.927 | 0.898 | 0.857 | 11.3s |
| auto-b62c0378 | 0.974 | 1.000 | 1.000 | 0.927 | 0.898 | 0.857 | 11.0s |