paper-with-me

Papers

Information-Theoretic Discrete Diffusion

2025-10-28 · Moongyu Jeon, Sangwoo Shin, Dongjae Jeon, Albert No arxiv

We present an information-theoretic framework for discrete diffusion models that yields principled estimators of log-likelihood using score-matching losses. Inspired by the I-MMSE identity for the Gaussian setup, we derive analogous results for the discrete setting. Specifically, we introduce the Information-Minimum Denoising Score Entropy (I-MDSE) relation, which links mutual information between data and its diffused version to the minimum denoising score entropy (DSE) loss. We extend this theory to masked diffusion and establish the Information-Minimum Denoising Cross-Entropy (I-MDCE) relation, connecting cross-entropy losses to mutual information in discrete masked processes. These results provide a time-integral decomposition of the log-likelihood of the data in terms of optimal score-based losses, showing that commonly used losses such as DSE and DCE are not merely variational bounds but tight and principled estimators of log-likelihood. The I-MDCE decomposition further enables practical extensions, including time-free formula, conditional likelihood estimation in prompt-response tasks, and coupled Monte Carlo estimation of likelihood ratios. Experiments on synthetic and real-world data confirm the accuracy, variance stability, and utility of our estimators. The code is publicly available at https://github.com/Dongjae0324/infodis.

📄 PDF Abstract BibTeX arXiv:2510.24088

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

ItDPDM: Information-Theoretic Discrete Poisson Diffusion Model

2025-05-08 · Sagnik Bhattacharya, Abhiram Gorle, Ahsan Bilal, Connor Ding 외

Generative modeling of non-negative, discrete data, such as symbolic music, remains challenging due to two persistent limitations in existing methods. Firstly, many approaches rely on modeling continuous embeddings, whic…

model

Information-Theoretic Lower Bounds for Recovery of Diffusion Network Structures

2016-01-28 · Keehwan Park, Jean Honorio

We study the information-theoretic lower bound of the sample complexity of the correct recovery of diffusion network structures. We introduce a discrete-time diffusion model based on the Independent Cascade model for whi…

Discrete State Diffusion Models: A Sample Complexity Perspective

2025-10-12 · Aadithya Srikanth, Mudit Gaur, Vaneet Aggarwal arxiv

Diffusion models have demonstrated remarkable performance in generating high-dimensional samples across domains such as vision, language, and the sciences. Although continuous-state diffusion models have been extensively…

Overcoming Dimensional Factorization Limits in Discrete Diffusion Models through Quantum Joint Distribution Learning

2025-05-08 · Chuangtao Chen, Qinglin Zhao, Mengchu Zhou, Zhimin He 외

This study explores quantum-enhanced discrete diffusion models to overcome classical limitations in learning high-dimensional distributions. We rigorously prove that classical discrete diffusion models, which calculate p…

Denoising

Coevolutionary Continuous Discrete Diffusion: Make Your Diffusion Language Model a Latent Reasoner

2025-10-03 · Cai Zhou, Chenxiao Yang, Yi Hu, Chenyu Wang 외 arxiv

Diffusion language models, especially masked discrete diffusion models, have achieved great success recently. While there are some theoretical and primary empirical results showing the advantages of latent reasoning with…