PT  - JOURNAL ARTICLE
AU  - Koo, Peter K.
AU  - Ploenzke, Matt
TI  - Improving Convolutional Network Interpretability with Exponential Activations
AID  - 10.1101/650804
DP  - 2019 Jan 01
TA  - bioRxiv
PG  - 650804
4099  - http://biorxiv.org/content/early/2019/05/27/650804.short
4100  - http://biorxiv.org/content/early/2019/05/27/650804.full
AB  - Deep convolutional networks trained on regulatory genomic sequences tend to learn distributed representations of sequence motifs across many first layer filters. This makes it challenging to decipher which features are biologically meaningful. Here we introduce the exponential activation that – when applied to first layer filters – leads to more interpretable representations of motifs, both visually and quantitatively, compared to rectified linear units. We demonstrate this on synthetic DNA sequences which have ground truth with various convolutional networks, and then show that this phenomenon holds on in vivo DNA sequences.