chr5tphr / zennit

Zennit is a high-level framework in Python using PyTorch for explaining/exploring neural networks using attribution methods like LRP.

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

Canonizers: BatchNorm loses bias relevance

chr5tphr opened this issue · comments

With the implementation of #185 allowing for gradient computation wrt. the parameters given an attribution with canonizers, the BatchNorms seem to be leaking attribution even with zero_params='bias' on ResNet18.