Is AI Robust Enough for Scientific Research?
Jun-Jie Zhang, Jiahao Song, Xiu-Cheng Wang, Fu-Peng Li, Zehan Liu, Jian-Nan Chen, Haoning Dang, Shiyao Wang, Yiyan Zhang, Jianhui Xu, Chunxiang Shi, Fei Wang, Long-Gang Pang, Nan Cheng, Weiwei Zhang, Duo Zhang, Deyu Meng
Unverified — Be the first to reproduce this paper.
ReproduceAbstract
We uncover a phenomenon largely overlooked by the scientific community utilizing AI: neural networks exhibit high susceptibility to minute perturbations, resulting in significant deviations in their outputs. Through an analysis of five diverse application areas -- weather forecasting, chemical energy and force calculations, fluid dynamics, quantum chromodynamics, and wireless communication -- we demonstrate that this vulnerability is a broad and general characteristic of AI systems. This revelation exposes a hidden risk in relying on neural networks for essential scientific computations, calling further studies on their reliability and security.