Extracting Automata from Recurrent Neural Networks Using Queries and Counterexamples [article]

Gail Weiss and Yoav Goldberg and Eran Yahav
2020 arXiv   pre-print
We present a novel algorithm that uses exact learning and abstraction to extract a deterministic finite automaton describing the state dynamics of a given trained RNN. We do this using Angluin's L* algorithm as a learner and the trained RNN as an oracle. Our technique efficiently extracts accurate automata from trained RNNs, even when the state vectors are large and require fine differentiation.
arXiv:1711.09576v4 fatcat:y57bs42rprbnlf6klol3crcrj4