PCLR Fold 3

Back to PCLR overview · Back to main index

Fold metrics

Metric	Value
Test accuracy	8.93%
Test F1 score	0.0938
Hierarchical loss	0.76378547
P-adic loss (total)	772.19138261
P-adic loss (mean)	0.84116708
Prime base	79
Training samples	3,662
Test samples	918
Trained at	2026-08-28T05:31:41+10:00

P-adic loss breakdown

Agreement	Count	Share	Cost per mistake	Total contribution
Exact match	82	8.93%	0.000000	0.000000
p^4	7	0.76%	0.000000	0.000000
p^3	33	3.59%	0.000002	0.000067
p^2	9	0.98%	0.000160	0.001442
p^1	15	1.63%	0.012658	0.189873
p^0	772	84.10%	1.000000	772.000000

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 79). Lower values indicate closer predictions in the taxonomy hierarchy. This metric is shared with the umllr model for comparison.