Is AI Robust Enough for Scientific Research?
Machine Learning
2024-12-24 v1 Computational Physics
Abstract
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.
Cite
@article{arxiv.2412.16234,
title = {Is AI Robust Enough for Scientific Research?},
author = {Jun-Jie Zhang and Jiahao Song and Xiu-Cheng Wang and Fu-Peng Li and Zehan Liu and Jian-Nan Chen and Haoning Dang and Shiyao Wang and Yiyan Zhang and Jianhui Xu and Chunxiang Shi and Fei Wang and Long-Gang Pang and Nan Cheng and Weiwei Zhang and Duo Zhang and Deyu Meng},
journal= {arXiv preprint arXiv:2412.16234},
year = {2024}
}
Comments
26 pages, 6 figures