Skip to yearly menu bar Skip to main content


Neural Network Attributions: A Causal Perspective

Aditya Chattopadhyay · Piyushi Manupriya · Anirban Sarkar · Vineeth N Balasubramanian

Pacific Ballroom #61

Keywords: [ Others ] [ Interpretability ] [ Causality ]


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.

Live content is unavailable. Log in and register to view live content