Weighted Automata Extraction from Recurrent Neural Networks via Regression on State Spaces

@inproceedings{Okudono2020WeightedAE,
  title={Weighted Automata Extraction from Recurrent Neural Networks via Regression on State Spaces},
  author={Takamasa Okudono and Masaki Waga and Taro Sekiyama and I. Hasuo},
  booktitle={AAAI},
  year={2020}
}
We present a method to extract a weighted finite automaton (WFA) from a recurrent neural network (RNN). Our algorithm is based on the WFA learning algorithm by Balle and Mohri, which is in turn an extension of Angluin's classic \lstar algorithm. Our technical novelty is in the use of \emph{regression} methods for the so-called equivalence queries, thus exploiting the internal state space of an RNN to prioritize counterexample candidates. This way we achieve a quantitative/weighted extension of… Expand
10 Citations
Property-Directed Verification of Recurrent Neural Networks
  • 2
  • PDF
Learning Deterministic Weighted Automata with Queries and Counterexamples
  • 11
  • PDF
Approximating probabilistic models as weighted finite automata
  • 2
  • PDF
A Formal Language Approach to Explaining RNNs
  • PDF
Genetic algorithm for the weight maximization problem on weighted automata
  • 1
  • PDF
A Passive Online Technique for Learning Hybrid Automata from Input/Output Traces
  • PDF

References

SHOWING 1-10 OF 54 REFERENCES
Extraction of rules from discrete-time recurrent neural networks
  • 197
  • PDF
Explaining Black Boxes on Sequential Data using Weighted Automata
  • 14
  • PDF
State-Regularized Recurrent Neural Networks
  • 14
  • PDF
Learning and Extracting Finite State Automata with Second-Order Recurrent Neural Networks
  • 473
Learning a class of large finite state machines with a recurrent neural network
  • 62
  • PDF
Spectral learning of weighted automata
  • 68
  • PDF
DyNet: The Dynamic Neural Network Toolkit
  • 329
  • PDF
...
1
2
3
4
5
...