Probabilistic Verification of Recurrent Neural Networks for Single and Multi-Agent Reinforcement Learning

ArXi:2605.14758v1 Announce Type: new History-dependent policies induced by recurrent neural networks (RNNs) rely on latent hidden state dynamics, making verification in partially observable reinforcement learning (RL) challenging. Existing RNN verification tools typically rely on restrictive modeling assumptions or coarse over-approximations of the hidden state space, which can lead to overly conservative or inconclusive results.