Zeroth-Order Sharpness-Aware Learning with Exponential Tilting
2510.16157v1
cs.LG, cs.CL, stat.ML
2025-10-22
Авторы:
Xuchen Gong, Tian Li
Abstract
Classic zeroth-order optimization approaches typically optimize for a
smoothed version of the original function, i.e., the expected objective under
randomly perturbed model parameters. This can be interpreted as encouraging the
loss values in the perturbation set to be small on average. Popular
sharpness-aware minimization (SAM) objectives, however, typically focus on the
largest loss within the neighborhood to arrive at flat minima more effectively.
In this work, we connect zeroth-order optimization (and its corresponding
objectives) with SAM approaches explicitly, through an exponential tilting
objective that provides a smooth transition between the average- and the
max-loss formulations. We explore new zeroth-order algorithms to solve a soft
SAM objective parameterized by a tilting parameter $t$. We provide precise
characterizations of the sharpness notions of the tilted SAM framework.
Practically, our approach can be used as a gradient-free and memory-efficient
alternative to SAM variants, and it achieves better generalization compared to
vanilla zeroth-order baselines on a wide range of downstream tasks, including
classification, multiple choice QA, and language generation.
Ссылки и действия
Дополнительные ресурсы: