# Gray-box Adversarial Testing for Control Systems with Machine Learning Component

arXiv: https://arxiv.org/abs/1812.11958 · HTML: https://mlsanity.com/paper/1812.11958/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 31 Dec 2018.

## Built on by (most cited first, 8 of 8)

- [Software Engineering for AI-Based Systems: A Survey](https://mlsanity.com/paper/2105.01984/) (2021)
- [How to Certify Machine Learning Based Safety-critical Systems? A Systematic Literature Review](https://mlsanity.com/paper/2107.12045/) (2021)
- [A Framework Based on Symbolic Regression Coupled with eXtended Physics-Informed Neural Networks for Gray-Box Learning of Equations of Motion from Data](https://mlsanity.com/paper/2305.10706/) (2023)
- [ACE: Self-Evolving LLM Coding Framework via Adversarial Unit Test Generation and Preference Optimization](https://mlsanity.com/paper/2605.16299/) (2026)
- [Falsification of Cyber-Physical Systems using Bayesian Optimization](https://mlsanity.com/paper/2209.06735/) (2022)
- [Scaling Learning based Policy Optimization for Temporal Logic Tasks by Controller Network Dropout](https://mlsanity.com/paper/2403.15826/) (2024)
- [Data-Driven Falsification of Cyber-Physical Systems](https://mlsanity.com/paper/2505.03863/) (2025)
- [Falsification-driven reinforcement learning for maritime motion planning](https://mlsanity.com/paper/2510.06970/) (2025)

## Abstract (arXiv)

Neural Networks (NN) have been proposed in the past as an effective means for both modeling and control of systems with very complex dynamics. However, despite the extensive research, NN-based controllers have not been adopted by the industry for safety critical systems. The primary reason is that systems with learning based controllers are notoriously hard to test and verify. Even harder is the analysis of such systems against system-level specifications. In this paper, we provide a gradient based method for searching the input space of a closed-loop control system in order to find adversarial samples against some system-level requirements. Our experimental results show that combined with randomized search, our method outperforms Simulated Annealing optimization.