On Tractable Representations of Binary Neural Networks
This addresses the need for interpretability and formal verification in neural networks, though it is incremental as it builds on existing compilation methods.
The paper tackles the problem of explaining and verifying binary neural networks by compiling their decision functions into tractable representations like OBDDs and SDDs, showing it is feasible to obtain compact SDD representations in experiments.
We consider the compilation of a binary neural network's decision function into tractable representations such as Ordered Binary Decision Diagrams (OBDDs) and Sentential Decision Diagrams (SDDs). Obtaining this function as an OBDD/SDD facilitates the explanation and formal verification of a neural network's behavior. First, we consider the task of verifying the robustness of a neural network, and show how we can compute the expected robustness of a neural network, given an OBDD/SDD representation of it. Next, we consider a more efficient approach for compiling neural networks, based on a pseudo-polynomial time algorithm for compiling a neuron. We then provide a case study in a handwritten digits dataset, highlighting how two neural networks trained from the same dataset can have very high accuracies, yet have very different levels of robustness. Finally, in experiments, we show that it is feasible to obtain compact representations of neural networks as SDDs.