PCLR Fold 2

Back to PCLR overview · Back to main index

Fold metrics

Metric	Value
Test accuracy	5.47%
Test F1 score	0.0598
Hierarchical loss	0.75387775
P-adic loss (total)	3719.62381115
P-adic loss (mean)	0.80441691
Prime base	83
Training samples	18,473
Test samples	4,624
Trained at	2026-10-13T05:39:24+11:00

P-adic loss breakdown

Agreement	Count	Share	Cost per mistake	Total contribution
Exact match	253	5.47%	0.000000	0.000000
p^4	9	0.19%	0.000000	0.000000
p^3	36	0.78%	0.000002	0.000063
p^2	396	8.56%	0.000145	0.057483
p^1	213	4.61%	0.012048	2.566265
p^0	3,717	80.38%	1.000000	3717.000000

Tag Rank vs Coefficient Magnitude

Tag rank vs max coefficient magnitude — Scatter plot showing the relationship between tag battle ranking and maximum absolute coefficient value across all taxonomies. Shows which tags have the strongest influence in the model.

About p-adic loss

P-adic loss measures the distance between predicted and true taxonomy using p-adic metric (base 83). Lower values indicate closer predictions in the taxonomy hierarchy. This metric is shared with the umllr model for comparison.