[Back]

R. Hasani, A. Amini, M. Lechner, F. Naser, R. Grosu, D. Rus:
"Response Characterization for Auditing Cell Dynamics in Long Short-term Memory Networks";
Talk: Workshop on Interpretability and Robustness in Audio, Speech, and Language (IRASL) at NIPS 2018, Montreal, Canada; 2018-12-02 - 2018-12-09; in: "Proceedings of the NIPS 2018 Interpretability and Robustness for Audio, Speech and Language Workshop", NIPS 2018, (2018).

In this paper, we introduce a novel method to interpret recurrent neural networks (RNNs), particularly long short-term memory networks (LSTMs) at the cellular level. We propose a systematic pipeline for interpreting individual hidden state dynamics within the network using response characterization methods. The ranked contribution of individual cells to the network's output is computed by analyzing a set of interpretable metrics of their decoupled step and sinusoidal responses. As a result, our method is able to uniquely identify neurons with insightful dynamics, quantify relationships between dynamical properties and test accuracy through ablation analysis, and interpret the impact of network capacity on a network's dynamical distribution. Finally, we demonstrate generalizability and scalability of our method by evaluating a series of different benchmark sequential datasets.