Verification of RNN-Based Neural Agent-Environment Systems

Authors

  • Michael E. Akintunde Imperial College London
  • Andreea Kevorchian Imperial College London
  • Alessio Lomuscio Imperial College London
  • Edoardo Pirovano Imperial College London

DOI:

https://doi.org/10.1609/aaai.v33i01.33016006

Abstract

We introduce agent-environment systems where the agent is stateful and executing a ReLU recurrent neural network. We define and study their verification problem by providing equivalences of recurrent and feed-forward neural networks on bounded execution traces. We give a sound and complete procedure for their verification against properties specified in a simplified version of LTL on bounded executions. We present an implementation and discuss the experimental results obtained.

AAAI-19 Proceedings Cover

Downloads

Published

2019-07-17

How to Cite

Akintunde, M. E., Kevorchian, A., Lomuscio, A., & Pirovano, E. (2019). Verification of RNN-Based Neural Agent-Environment Systems. Proceedings of the AAAI Conference on Artificial Intelligence, 33(01), 6006-6013. https://doi.org/10.1609/aaai.v33i01.33016006

Issue

Vol. 33 No. 01: AAAI-19, IAAI-19, EAAI-20

Section

AAAI Technical Track: Multiagent Systems