PCLR Fold 3

Back to PCLR overview · Back to main index

Fold metrics

Metric	Value
Test accuracy	4.96%
Test F1 score	0.0485
Hierarchical loss	0.76058152
P-adic loss (total)	2109.18846995
P-adic loss (mean)	0.76949598
Prime base	79
Training samples	10,818
Test samples	2,741
Trained at	2026-05-25T05:31:50+10:00

P-adic loss breakdown

Agreement	Count	Share	Cost per mistake	Total contribution
Exact match	136	4.96%	0.000000	0.000000
p^4	3	0.11%	0.000000	0.000000
p^3	98	3.58%	0.000002	0.000199
p^2	227	8.28%	0.000160	0.036372
p^1	170	6.20%	0.012658	2.151899
p^0	2,107	76.87%	1.000000	2107.000000

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.