# Forget Sharpness: Perturbed Forgetting of Model Biases Within SAM Dynamics

Source: [Forget Sharpness: Perturbed Forgetting of Model Biases Within SAM Dynamics](<https://www.sota2.com/research/paper/68fa73971e209131cea5d6e3>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-28T13:57:07.240Z

Authors: Ankit Vani, Frederick Tung, Gabriel L. Oliveira, Hossein Sharifi-Noghabi

Published: 2024-06-10T18:02:48.000Z

## Abstract

Despite attaining high empirical generalization, the sharpness of models trained with sharpness-aware minimization (SAM) do not always correlate with generalization error. Instead of viewing SAM as minimizing sharpness to improve generalization, our paper considers a new perspective based on SAM's training dynamics. We propose that perturbations in SAM perform perturbed forgetting, where they discard undesirable model biases to exhibit learning signals that generalize better. We relate our notion of forgetting to the information bottleneck principle, use it to explain observations like the better generalization of smaller perturbation batches, and show that perturbed forgetting can exhibit a stronger correlation with generalization than flatness. While standard SAM targets model biases exposed by the steepest ascent directions, we propose a new perturbation that targets biases exposed through the model's outputs. Our output bias forgetting perturbations outperform standard SAM, GSAM, and ASAM on ImageNet, robustness benchmarks, and transfer to CIFAR-{10,100}, while sometimes converging to sharper regions. Our results suggest that the benefits of SAM can be explained by alternative mechanistic principles that do not require flatness of the loss surface.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2406.06700>)

[PDF](<https://arxiv.org/pdf/2406.06700.pdf>)
