PCNN Fold 3

Back to PCNN overview · Back to main index

Fold metrics

Metric	Value
Test accuracy	13.83%
Test F1 score	0.0933
Hierarchical loss	0.76599230
P-adic loss (total)	754.17804000
P-adic loss (mean)	0.82154470
Prime base	79
Hidden layer size	27
Max tags	32
Training samples	3,662
Test samples	918

P-adic loss breakdown

Agreement	Count	Share	Cost per mistake	Total contribution
Exact match	127	13.83%	0.000000	0.000000
p^4	3	0.33%	0.000000	0.000000
p^3	15	1.63%	0.000002	0.000030
p^2	5	0.54%	0.000160	0.000801
p^1	14	1.53%	0.012658	0.177215
p^0	754	82.14%	1.000000	754.000000

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.