Optimization and Abstraction: A Synergistic Approach for Analyzing Neural Network Robustness
In recent years, the notion of local robustness (or robustness for short) has emerged as a desirable property of deep neural networks. Intuitively, robustness means that small perturbations to an input do not cause the network to perform misclassifications. In this paper, we present a novel algorithm for verifying robustness properties of neural networks. Our method synergistically combines gradient-based optimization methods for counterexample search with abstraction-based proof search to obtain a sound and (δ -)complete decision procedure. Our method also employs a data-driven approach to learn a verification policy that guides abstract interpretation during proof search. We have implemented the proposed approach in a tool called Charon and experimentally evaluated it on hundreds of benchmarks. Our experiments show that the proposed approach significantly outperforms three state-of-the-art tools, namely AI^2, Reluplex, and Reluval.
Tue 25 Jun Times are displayed in time zone: Tijuana, Baja California change
14:00 - 15:30: Reasoning and Optimizing ML ModelsPLDI Research Papers at 224AB Chair(s): Martin MaasGoogle | |||
14:00 - 14:20 Talk | An Inductive Synthesis Framework for Verifiable Reinforcement Learning PLDI Research Papers He ZhuRutgers University, USA, Zikang XiongPurdue University, Stephen Magill, Suresh JagannathanPurdue University Media Attached | ||
14:20 - 14:40 Talk | Programming Support for Autonomizing Software PLDI Research Papers Wen-Chuan LeePurdue University, Peng LiuPurdue University, Yingqi LiuPurdue University, USA, Shiqing MaPurdue University, USA, Xiangyu ZhangPurdue University | ||
14:40 - 15:00 Talk | Wootz: A Compiler-Based Framework for Fast CNN Pruning via Composability PLDI Research Papers Hui GuanNorth Carolina State University, Xipeng ShenNorth Carolina State University, Seung-Hwan LimOak Ridge National Laboratory, USA Media Attached File Attached | ||
15:00 - 15:20 Talk | Optimization and Abstraction: A Synergistic Approach for Analyzing Neural Network Robustness PLDI Research Papers Greg AndersonUniversity of Texas at Austin, USA, Shankara PailoorUniversity of Texas at Austin, USA, Isil DilligUT Austin, Swarat ChaudhuriRice University Media Attached |