Convergence Guarantees for Time-Inhomogeneous Uniform-Rate Discrete Diffusion Models
Entropy · 11.06.2026 · 28 (6) · 675
DOI: 10.3390/e28060675 ↗Kurz gefasst
Fachartikel in der Zeitschrift Entropy, online erschienen am 11.06.2026. Verfasst von Yuchen Liang und 3 weiteren Autor:innen. Beteiligte Einrichtung: Luddy School of Informatics, Computing, and Engineering, Indiana University Indianapolis, Indianapolis, IN 46202, USA. Die Arbeit nennt 27 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: U.S. National Science Foundation, DEVCOM Army Research Laboratory.
Abstract (Originalsprache)
Discrete diffusion models have become an important class of generative models for categorical data, yet their theoretical understanding remains largely limited to time-homogeneous noise schedules. In this work, we study uniform-rate discrete diffusion models with time-inhomogeneous continuous-time Markov chain forward processes. We establish convergence guarantees for practical reverse-time samplers by directly controlling the total variation distance, avoiding the indirect route of first bounding KL divergence and then applying Pinsker’s inequality. Our analysis decomposes the sampling error into initialization, score-estimation, discretization, and early-stopping errors, and explicitly characterizes how each term depends on the accumulated noise, the local noise rate, and the smoothness of the noise schedule. Under suitable regularity conditions on the noise schedule, we further derive step-complexity guarantees that match the order of existing results for homogeneous samplers.
Schlagwörter
Stichwörter der Autor:innen
Autor:innen und Institutionen
- Luddy School of Informatics, Computing, and Engineering, Indiana University Indianapolis, Indianapolis, IN 46202, USA.
- Department of Electrical and Computer Engineering, UC Davis, Davis, CA 95616, USA.
- Department of Electrical and Computer Engineering, The Ohio State University, Columbus, OH 43210, USA.
ORCID-Profile: Yuchen Liang
Förderung
- U.S. National Science Foundation
- DEVCOM Army Research Laboratory
Zitieren
APA
Liang, Y., Lai, L., Shroff, N., & Liang, Y. (2026). Convergence Guarantees for Time-Inhomogeneous Uniform-Rate Discrete Diffusion Models. Entropy, 28(6), 675. https://doi.org/10.3390/e28060675
Vancouver
Liang Y, Lai L, Shroff N, Liang Y. Convergence Guarantees for Time-Inhomogeneous Uniform-Rate Discrete Diffusion Models. Entropy. 2026;28(6):675. doi:10.3390/e28060675
BibTeX
@article{Liang2026,
author = {Liang, Yuchen and Lai, Lifeng and Shroff, Ness and Liang, Yingbin},
title = {Convergence Guarantees for Time-Inhomogeneous Uniform-Rate Discrete Diffusion Models},
journal = {Entropy},
year = {2026},
volume = {28},
number = {6},
pages = {675},
doi = {10.3390/e28060675}
}
EUVIMED