NeVer 2.0: Learning, Verification and Repair of Deep Neural Networks
This work aims to provide an integrated tool for learning, verification, and repair of deep neural networks, which is a significant problem for researchers and practitioners working on reliable AI systems.
This paper introduces NeVer 2.0, a prototype system for automated synthesis and analysis of deep neural networks. It integrates state-of-the-art learning frameworks with verification algorithms to address scalability and enable repair of faulty deep networks.
In this work, we present an early prototype of NeVer 2.0, a new system for automated synthesis and analysis of deep neural networks.NeVer 2.0borrows its design philosophy from NeVer, the first package that integrated learning, automated verification and repair of (shallow) neural networks in a single tool. The goal of NeVer 2.0 is to provide a similar integration for deep networks by leveraging a selection of state-of-the-art learning frameworks and integrating them with verification algorithms to ease the scalability challenge and make repair of faulty networks possible.