Regularizing Differentiable Architecture Search with Smooth Activation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Zhou, Yanlin, El-Khamy, Mostafa, Song, Kee-Bong
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866909588942487552
author Zhou, Yanlin
El-Khamy, Mostafa
Song, Kee-Bong
author_facet Zhou, Yanlin
El-Khamy, Mostafa
Song, Kee-Bong
contents Differentiable Architecture Search (DARTS) is an efficient Neural Architecture Search (NAS) method but suffers from robustness, generalization, and discrepancy issues. Many efforts have been made towards the performance collapse issue caused by skip dominance with various regularization techniques towards operation weights, path weights, noise injection, and super-network redesign. It had become questionable at a certain point if there could exist a better and more elegant way to retract the search to its intended goal -- NAS is a selection problem. In this paper, we undertake a simple but effective approach, named Smooth Activation DARTS (SA-DARTS), to overcome skip dominance and discretization discrepancy challenges. By leveraging a smooth activation function on architecture weights as an auxiliary loss, our SA-DARTS mitigates the unfair advantage of weight-free operations, converging to fanned-out architecture weight values, and can recover the search process from skip-dominance initialization. Through theoretical and empirical analysis, we demonstrate that the SA-DARTS can yield new state-of-the-art (SOTA) results on NAS-Bench-201, classification, and super-resolution. Further, we show that SA-DARTS can help improve the performance of SOTA models with fewer parameters, such as Information Multi-distillation Network on the super-resolution task.
format Preprint
id arxiv_https___arxiv_org_abs_2504_16306
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Regularizing Differentiable Architecture Search with Smooth Activation
Zhou, Yanlin
El-Khamy, Mostafa
Song, Kee-Bong
Neural and Evolutionary Computing
Computer Vision and Pattern Recognition
Differentiable Architecture Search (DARTS) is an efficient Neural Architecture Search (NAS) method but suffers from robustness, generalization, and discrepancy issues. Many efforts have been made towards the performance collapse issue caused by skip dominance with various regularization techniques towards operation weights, path weights, noise injection, and super-network redesign. It had become questionable at a certain point if there could exist a better and more elegant way to retract the search to its intended goal -- NAS is a selection problem. In this paper, we undertake a simple but effective approach, named Smooth Activation DARTS (SA-DARTS), to overcome skip dominance and discretization discrepancy challenges. By leveraging a smooth activation function on architecture weights as an auxiliary loss, our SA-DARTS mitigates the unfair advantage of weight-free operations, converging to fanned-out architecture weight values, and can recover the search process from skip-dominance initialization. Through theoretical and empirical analysis, we demonstrate that the SA-DARTS can yield new state-of-the-art (SOTA) results on NAS-Bench-201, classification, and super-resolution. Further, we show that SA-DARTS can help improve the performance of SOTA models with fewer parameters, such as Information Multi-distillation Network on the super-resolution task.
title Regularizing Differentiable Architecture Search with Smooth Activation
topic Neural and Evolutionary Computing
Computer Vision and Pattern Recognition
url https://arxiv.org/abs/2504.16306