Extracting Weighted Finite Automata from Recurrent Neural Networks for Natural Languages

TL;DR

This paper introduces a scalable method for extracting weighted finite automata from RNNs in natural language processing, improving extraction precision by addressing transition sparsity and enhancing context-awareness.

Contribution

It proposes a novel transition rule extraction approach with data augmentation tactics, significantly improving the interpretability of RNNs for natural language tasks.

Findings

Outperforms existing methods in extraction precision

Effective in capturing dynamic behaviors of RNNs

Scalable to natural language processing models

Abstract

Recurrent Neural Networks (RNNs) have achieved tremendous success in sequential data processing. However, it is quite challenging to interpret and verify RNNs' behaviors directly. To this end, many efforts have been made to extract finite automata from RNNs. Existing approaches such as exact learning are effective in extracting finite-state models to characterize the state dynamics of RNNs for formal languages, but are limited in the scalability to process natural languages. Compositional approaches that are scablable to natural languages fall short in extraction precision. In this paper, we identify the transition sparsity problem that heavily impacts the extraction precision. To address this problem, we propose a transition rule extraction approach, which is scalable to natural language processing models and effective in improving extraction precision. Specifically, we propose an empirical method to complement the missing rules in the transition diagram. In addition, we further adjust the transition matrices to enhance the context-aware ability of the extracted weighted finite automaton (WFA). Finally, we propose two data augmentation tactics to track more dynamic behaviors of the target RNN. Experiments on two popular natural language datasets show that our method can extract WFA from RNN for natural language processing with better precision than existing approaches. Our code is available at https://github.com/weizeming/Extract_WFA_from_RNN_for_NL.