PCNN Fold 2

Back to PCNN overview · Back to main index

Fold metrics

Metric	Value
Test accuracy	14.07%
Test F1 score	0.1034
Hierarchical loss	0.78479190
P-adic loss (total)	1224.57860000
P-adic loss (mean)	0.67026746
Prime base	79
Hidden layer size	27
Max tags	32
Training samples	7,283
Test samples	1,827

P-adic loss breakdown

Agreement	Count	Share	Cost per mistake	Total contribution
Exact match	257	14.07%	0.000000	0.000000
p^4	4	0.22%	0.000000	0.000000
p^3	86	4.71%	0.000002	0.000174
p^2	134	7.33%	0.000160	0.021471
p^1	123	6.73%	0.012658	1.556962
p^0	1,223	66.94%	1.000000	1223.000000

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