Publications

Provable preimage under-approximation for neural networks

Published in Proceedings of the 30th International Conference on Tools and Algorithms for the Construction and Analysis of Systems, 2024

Recommended citation: Xiyue Zhang, Benjie Wang, Marta Kwiatkowska . "Provable Preimage Under-Approximation for Neural Networks." Proceedings of the 30th International Conference on Tools and Algorithms for the Construction and Analysis of Systems. TACAS 2024. https://link.springer.com/chapter/10.1007/978-3-031-57256-2_1

Weighted Automata Extraction and Explanation of Recurrent Neural Networks for Natural Language Tasks

Published in Journal of Logical and Algebraic Methods in Programming, 2023

Recommended citation: Zeming Wei, Xiyue Zhang, Yihao Zhang, Meng Sun. "Weighted Automata Extraction and Explanation of Recurrent Neural Networks for Natural Language Tasks." Journal of Logical and Algebraic Methods in Programming. Volume 136, https://doi.org/10.1016/j.jlamp.2023.100907. https://authors.elsevier.com/c/1hnCS8MrKMbfOd

kProp: Multi-Neuron Relaxation Method for Neural Network Robustness Verification

Published in Proceedings of the 10th International Conference on Fundamentals of Software Engineering, 2022

Recommended citation: Xiaoyong Xue, Xiyue Zhang, Meng Sun. "kProp: Multi-Neuron Relaxation Method for Neural Network Robustness Verification." Proceedings of the 10th International Conference on Fundamentals of Software Engineering. FSEN 2023.

Towards Characterizing Adversarial Defects of Deep Learning Software from the Lens of Uncertainty

Published in Proceedings of the 42nd IEEE/ACM International Conference on Software Engineering, 2020

Recommended citation: Xiyue Zhang, Xiaofei Xie, Lei Ma, Xiaoning Du, Qiang Hu, Yang Liu, Jianjun Zhao, Meng Sun. "Towards Characterizing Adversarial Defects of Deep Learning Software from the Lens of Uncertainty." Proceedings of the 42nd IEEE/ACM International Conference on Software Engineering. ICSE 2020. https://arxiv.org/abs/2004.11573

Using Recurrent Neural Network to Predict Tactics for Proving Component Connector Properties in Coq

Published in Proceedings of International Symposium on Theoretical Aspects of Software Engineering, 2019

Recommended citation: Xiyue Zhang, Yi Li, Weijiang Hong, Meng Sun. "Using Recurrent Neural Network to Predict Tactics for Proving Component Connector Properties in Coq." Proceedings of International Symposium on Theoretical Aspects of Software Engineering. TASE 2019. https://ieeexplore.ieee.org/document/8914103