Neural Architecture Search Finds Robust Models by Knowledge Distillation

Utkarsh Nath, Yancheng Wang, Yingzhen Yang
Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence, PMLR 244:2700-2715, 2024.

Abstract

Despite their superior performance, Deep Neural Networks (DNNs) are often vulnerable to adversarial attacks. Neural Architecture Search (NAS), a method for automatically designing the architectures of DNNs, has shown remarkable performance across various machine learning applications. However, the adversarial robustness of architectures learned by NAS against adversarial threats remains under-explored. By integrating a robust teacher, we examine whether NAS can yield a robust neural architecture by inheriting robustness from the teacher. In this paper, we propose Robust Neural Architecture Search by Cross-Layer Knowledge Distillation (RNAS-CL), a novel NAS algorithm that enhances the robustness of architectures learned by NAS through employing cross-layer knowledge distillation from a robust teacher. Distinct from previous knowledge distillation approaches that only align student-teacher outputs at the final layer, RNAS-CL dynamically searches for the optimal teacher layer to guide each student layer. Our experimental findings validate the effectiveness of RNAS-CL, demonstrating that it can generate both compact and adversarially robust neural architectures. Our results pave the way for developing new strategies for compact and robust neural architecture design applicable across various fields. The code of RNAS-CL is available at \url{https://github.com/Statistical-Deep-Learning/RNAS-CL}.

Cite this Paper

BibTeX
@InProceedings{pmlr-v244-nath24a, title = {Neural Architecture Search Finds Robust Models by Knowledge Distillation}, author = {Nath, Utkarsh and Wang, Yancheng and Yang, Yingzhen}, booktitle = {Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence}, pages = {2700--2715}, year = {2024}, editor = {Kiyavash, Negar and Mooij, Joris M.}, volume = {244}, series = {Proceedings of Machine Learning Research}, month = {15--19 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v244/main/assets/nath24a/nath24a.pdf}, url = {https://proceedings.mlr.press/v244/nath24a.html}, abstract = {Despite their superior performance, Deep Neural Networks (DNNs) are often vulnerable to adversarial attacks. Neural Architecture Search (NAS), a method for automatically designing the architectures of DNNs, has shown remarkable performance across various machine learning applications. However, the adversarial robustness of architectures learned by NAS against adversarial threats remains under-explored. By integrating a robust teacher, we examine whether NAS can yield a robust neural architecture by inheriting robustness from the teacher. In this paper, we propose Robust Neural Architecture Search by Cross-Layer Knowledge Distillation (RNAS-CL), a novel NAS algorithm that enhances the robustness of architectures learned by NAS through employing cross-layer knowledge distillation from a robust teacher. Distinct from previous knowledge distillation approaches that only align student-teacher outputs at the final layer, RNAS-CL dynamically searches for the optimal teacher layer to guide each student layer. Our experimental findings validate the effectiveness of RNAS-CL, demonstrating that it can generate both compact and adversarially robust neural architectures. Our results pave the way for developing new strategies for compact and robust neural architecture design applicable across various fields. The code of RNAS-CL is available at \url{https://github.com/Statistical-Deep-Learning/RNAS-CL}.} }
Endnote
%0 Conference Paper %T Neural Architecture Search Finds Robust Models by Knowledge Distillation %A Utkarsh Nath %A Yancheng Wang %A Yingzhen Yang %B Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2024 %E Negar Kiyavash %E Joris M. Mooij %F pmlr-v244-nath24a %I PMLR %P 2700--2715 %U https://proceedings.mlr.press/v244/nath24a.html %V 244 %X Despite their superior performance, Deep Neural Networks (DNNs) are often vulnerable to adversarial attacks. Neural Architecture Search (NAS), a method for automatically designing the architectures of DNNs, has shown remarkable performance across various machine learning applications. However, the adversarial robustness of architectures learned by NAS against adversarial threats remains under-explored. By integrating a robust teacher, we examine whether NAS can yield a robust neural architecture by inheriting robustness from the teacher. In this paper, we propose Robust Neural Architecture Search by Cross-Layer Knowledge Distillation (RNAS-CL), a novel NAS algorithm that enhances the robustness of architectures learned by NAS through employing cross-layer knowledge distillation from a robust teacher. Distinct from previous knowledge distillation approaches that only align student-teacher outputs at the final layer, RNAS-CL dynamically searches for the optimal teacher layer to guide each student layer. Our experimental findings validate the effectiveness of RNAS-CL, demonstrating that it can generate both compact and adversarially robust neural architectures. Our results pave the way for developing new strategies for compact and robust neural architecture design applicable across various fields. The code of RNAS-CL is available at \url{https://github.com/Statistical-Deep-Learning/RNAS-CL}.
APA
Nath, U., Wang, Y. & Yang, Y.. (2024). Neural Architecture Search Finds Robust Models by Knowledge Distillation. Proceedings of the Fortieth Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 244:2700-2715 Available from https://proceedings.mlr.press/v244/nath24a.html.

Related Material