Property-Directed Verification of Recurrent Neural Networks

09/22/2020
by   Igor Khmelnitsky, et al.
24

This paper presents a property-directed approach to verifying recurrent neural networks (RNNs). To this end, we learn a deterministic finite automaton as a surrogate model from a given RNN using active automata learning. This model may then be analyzed using model checking as verification technique. The term property-directed reflects the idea that our procedure is guided and controlled by the given property rather than performing the two steps separately. We show that this not only allows us to discover small counterexamples fast, but also to generalize them by pumping towards faulty flows hinting at the underlying error in the RNN.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
10/09/2019

Generalized Property-Directed Reachability for Hybrid Systems

Generalized property-directed reachability (GPDR) belongs to the family ...
research
11/04/2021

Small model property reflects in games and automata

Small model property is an important property that implies decidability....
research
11/14/2018

Verification of Recurrent Neural Networks Through Rule Extraction

The verification problem for neural networks is verifying whether a neur...
research
01/20/2021

Synthesizing Context-free Grammars from Recurrent Neural Networks (Extended Version)

We present an algorithm for extracting a subclass of the context free gr...
research
04/06/2020

Verifying Recurrent Neural Networks using Invariant Inference

Deep neural networks are revolutionizing the way complex systems are dev...
research
06/12/2020

A Formal Language Approach to Explaining RNNs

This paper presents LEXR, a framework for explaining the decision making...
research
04/23/2020

Bayesian Verification of Chemical Reaction Networks

We present a data-driven verification approach that determines whether o...

Please sign up or login with your details

Forgot password? Click here to reset