Selective Underfitting in Diffusion Models

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Song, Kiwhan, Kim, Jaeyeon, Chen, Sitan, Du, Yilun, Kakade, Sham, Sitzmann, Vincent
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866908572892266496
author Song, Kiwhan
Kim, Jaeyeon
Chen, Sitan
Du, Yilun
Kakade, Sham
Sitzmann, Vincent
author_facet Song, Kiwhan
Kim, Jaeyeon
Chen, Sitan
Du, Yilun
Kakade, Sham
Sitzmann, Vincent
contents Diffusion models have emerged as the principal paradigm for generative modeling across various domains. During training, they learn the score function, which in turn is used to generate samples at inference. They raise a basic yet unsolved question: which score do they actually learn? In principle, a diffusion model that matches the empirical score in the entire data space would simply reproduce the training data, failing to generate novel samples. Recent work addresses this question by arguing that diffusion models underfit the empirical score due to training-time inductive biases. In this work, we refine this perspective, introducing the notion of selective underfitting: instead of underfitting the score everywhere, better diffusion models more accurately approximate the score in certain regions of input space, while underfitting it in others. We characterize these regions and design empirical interventions to validate our perspective. Our results establish that selective underfitting is essential for understanding diffusion models, yielding new, testable insights into their generalization and generative performance.
format Preprint
id arxiv_https___arxiv_org_abs_2510_01378
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Selective Underfitting in Diffusion Models
Song, Kiwhan
Kim, Jaeyeon
Chen, Sitan
Du, Yilun
Kakade, Sham
Sitzmann, Vincent
Machine Learning
Diffusion models have emerged as the principal paradigm for generative modeling across various domains. During training, they learn the score function, which in turn is used to generate samples at inference. They raise a basic yet unsolved question: which score do they actually learn? In principle, a diffusion model that matches the empirical score in the entire data space would simply reproduce the training data, failing to generate novel samples. Recent work addresses this question by arguing that diffusion models underfit the empirical score due to training-time inductive biases. In this work, we refine this perspective, introducing the notion of selective underfitting: instead of underfitting the score everywhere, better diffusion models more accurately approximate the score in certain regions of input space, while underfitting it in others. We characterize these regions and design empirical interventions to validate our perspective. Our results establish that selective underfitting is essential for understanding diffusion models, yielding new, testable insights into their generalization and generative performance.
title Selective Underfitting in Diffusion Models
topic Machine Learning
url https://arxiv.org/abs/2510.01378