Code for the paper, Neural Network Attributions: A Causal Perspective (ICML 2019).

View the Project on GitHub Piyushi-0/ACE

Neural Network Attributions: A Causal Perspective

Aditya Chattopadhyay1  Piyushi Manupriya2  Anirban Sarkar2  Vineeth N Balasubramanian2

1Johns Hopkins University     2IIT Hyderabad

In ICML 2019

Paper | Code | Slides | Citations | Media Coverage


We propose a new attribution method for neural networks developed using first principles of causality (to the best of our knowledge, the first such). The neural network architecture is viewed as a Structural Causal Model, and a methodology to compute the causal effect of each feature on the output is presented. With reasonable assumptions on the causal structure of the input data,we propose algorithms to efficiently compute the causal effects, as well as scale the approach to data with large dimensionality. We also show how this method can be used for recurrent neural networks.We report experimental results on both simulated and real datasets showcasing the promise and usefulness of the proposed algorithm.

Paper on Arxiv

ICML Presentation

Some results

ACE on NASA dataset ACE on MNIST dataset


Aditya Chattopadhyay, Piyushi Manupriya, Anirban Sarkar, Vineeth N Balasubramanian. "Neural Network Attributions: A Causal Perspective", in International Conference on Machine Learning (ICML), 2019.


We are grateful to the Ministry of Human Resource Development, India; Department of Science and Technology, India;as well as Honeywell India for the financial support of this project through the UAY program. We thank the anonymous reviewers for their valuable feedback that helped improvethe presentation of this work.