# Approaching Deep Learning through the Spectral Dynamics of Weights

Source: [Approaching Deep Learning through the Spectral Dynamics of Weights](<https://www.sota2.com/research/paper/68fa748f1e209131cea627e3>)

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

Data updated: 2025-10-26T03:12:36.753Z

Authors: David Yunis, Kumar Kshitij Patel, Samuel Wheeler, Pedro Savarese, Gal Vardi, Karen Livescu, Michael Maire, Matthew R. Walter

Published: 2024-08-21T17:48:01.000Z

## Abstract

We propose an empirical approach centered on the spectral dynamics of weights -- the behavior of singular values and vectors during optimization -- to unify and clarify several phenomena in deep learning. We identify a consistent bias in optimization across various experiments, from small-scale \`\`grokking'' to large-scale tasks like image classification with ConvNets, image generation with UNets, speech recognition with LSTMs, and language modeling with Transformers. We also demonstrate that weight decay enhances this bias beyond its role as a norm regularizer, even in practical systems. Moreover, we show that these spectral dynamics distinguish memorizing networks from generalizing ones, offering a novel perspective on this longstanding conundrum. Additionally, we leverage spectral dynamics to explore the emergence of well-performing sparse subnetworks (lottery tickets) and the structure of the loss surface through linear mode connectivity. Our findings suggest that spectral dynamics provide a coherent framework to better understand the behavior of neural networks across diverse settings.

## 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/2408.11804>)

[PDF](<https://arxiv.org/pdf/2408.11804.pdf>)
