PCNN Fold 0

Back to PCNN overview · Back to main index

Fold metrics

Metric	Value
Test accuracy	11.81%
Test F1 score	0.0931
Hierarchical loss	0.78988740
P-adic loss (total)	2731.58740000
P-adic loss (mean)	0.59942670
Prime base	83
Hidden layer size	27
Max tags	32
Training samples	18,540
Test samples	4,557

P-adic loss breakdown

Agreement	Count	Share	Cost per mistake	Total contribution
Exact match	538	11.81%	0.000000	0.000000
p^3	196	4.30%	0.000002	0.000343
p^2	558	12.24%	0.000145	0.080999
p^1	540	11.85%	0.012048	6.506024
p^0	2,725	59.80%	1.000000	2725.000000

Tag Rank vs First-Layer Weight Magnitude

Tag rank vs max first-layer weight magnitude — Scatter plot showing the relationship between tag battle ranking and maximum absolute first-layer weight value across all hidden units. Shows which input features contribute most to the parameter constrained neural network's hidden representations.

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.