REGLO: Provable Neural Network Repair for Global Robustness Properties

Feisi Fu, Zhilu Wang, Weichao Zhou, Yixuan Wang, Jiameng Fan, Chao Huang, Qi Zhu, Xin Chen, Wenchao Li

Abstract

We present REGLO, a novel methodology for repairing pretrained neural networks to satisfy global robustness and individual fairness properties. A neural network is said to be globally robust with respect to a given input region if and only if all the input points in the region are locally robust. This notion of global robustness also captures the notion of individual fairness as a special case. We prove that any counterexample to a global robustness property must exhibit a corresponding large gradient. For ReLU networks, this result allows us to efficiently identify the linear regions that violate a given global robustness property. By formulating and solving a suitable robust convex optimization problem, REGLO then computes a minimal weight change that will provably repair these violating linear regions.
BibTeX
@inproceedings{fu2024reglo,
  author = {Feisi Fu and Zhilu Wang and Weichao Zhou and Yixuan Wang and Jiameng Fan and Chao Huang and Qi Zhu and Xin Chen and Wenchao Li},
  title = {{REGLO: Provable Neural Network Repair for Global Robustness Properties}},
  booktitle = {AAAI Conference on Artificial Intelligence},
  year = {2024},
  doi = {10.1609/aaai.v38i11.29094},
}
Figure