DeepAI AI Chat
Log In Sign Up

Attribution Preservation in Network Compression for Reliable Network Interpretation

by   Geondo Park, et al.

Neural networks embedded in safety-sensitive applications such as self-driving cars and wearable health monitors rely on two important techniques: input attribution for hindsight analysis and network compression to reduce its size for edge-computing. In this paper, we show that these seemingly unrelated techniques conflict with each other as network compression deforms the produced attributions, which could lead to dire consequences for mission-critical applications. This phenomenon arises due to the fact that conventional network compression methods only preserve the predictions of the network while ignoring the quality of the attributions. To combat the attribution inconsistency problem, we present a framework that can preserve the attributions while compressing a network. By employing the Weighted Collapsed Attribution Matching regularizer, we match the attribution maps of the network being compressed to its pre-compression former self. We demonstrate the effectiveness of our algorithm both quantitatively and qualitatively on diverse compression methods.


page 2

page 8

page 16

page 17


Discriminative Attribution from Counterfactuals

We present a method for neural network interpretability by combining fea...

Spatio-Temporal Perturbations for Video Attribution

The attribution method provides a direction for interpreting opaque neur...

A Self-contained Analysis of the Lempel-Ziv Compression Algorithm

This article gives a self-contained analysis of the performance of the L...

Robust Attribution Regularization

An emerging problem in trustworthy machine learning is to train models t...

Generating Attribution Maps with Disentangled Masked Backpropagation

Attribution map visualization has arisen as one of the most effective te...

An agent-driven semantical identifier using radial basis neural networks and reinforcement learning

Due to the huge availability of documents in digital form, and the decep...

Code Repositories


Attribution Preservation in Network Compression for Reliable Network Interpretation, NeurIPS 2020

view repo