Towards Interpreting Recurrent Neural Networks through Probabilistic Abstraction

Abstract

Neural networks are becoming a popular tool for solving many real-world problems such as object recognition and machine translation, thanks to its exceptional performance as an end-to-end solution. However, neural networks are complex black-box models, which hinders humans from interpreting and consequently trusting them in making critical decisions. Towards interpreting neural networks, several approaches have been proposed to extract simple deterministic models from neural networks. The results are not encouraging (e.g., low accuracy and limited scalability), fundamentally due to the limited expressiveness of such simple models. In this work, we propose an approach to extract probabilistic automata for interpreting an important class of neural networks, i.e., recurrent neural networks. Our work distinguishes itself from existing approaches in two important ways. One is that probability is used to compensate for the loss of expressiveness. This is inspired by the observation that human reasoning is often `probabilistic'. The other is that we adaptively identify the right level of abstraction so that a simple model is extracted in a request-specific way. We conduct experiments on several real-world datasets using state-of-the-art architectures including GRU and LSTM. The result shows that our approach significantly improves existing approaches in terms of accuracy or scalability. Lastly, we demonstrate the usefulness of the extracted models through detecting adversarial texts.

Accepted by ASE 2020

References in corpus (10)

• Efficient Estimation of Word Representations in Vector Space
• Learning Phrase Representations using RNN Encoder-Decoder for Statistical Machine Translation
• End to End Learning for Self-Driving Cars
• DeepXplore: Automated Whitebox Testing of Deep Learning Systems
• Targeted Backdoor Attacks on Deep Learning Systems Using Data Poisoning
• Seeing stars: Exploiting class relationships for sentiment categorization with respect to rating scales
• Extracting Automata from Recurrent Neural Networks Using Queries and Counterexamples
• HotFlip: White-Box Adversarial Examples for Text Classification
• A Survey of Safety and Trustworthiness of Deep Neural Networks: Verification, Testing, Adversarial Attack and Defence, and Interpretability
• Explaining Black Boxes on Sequential Data using Weighted Automata