Simplex Relaxation for Discrete Diffusion

• URL: http://arxiv.org/abs/2608.10615v1
• Date: Tue, 11 Aug 2026 07:58:17 GMT
• Title: Simplex Relaxation for Discrete Diffusion
• Abstract summary: We introduce Simplax, an exact Dirichlet--categorical augmentation that couples each corrupted categorical state with an auxiliary simplex-valued variable.<n>Simplax improves the generative perplexity--entropy tradeoff on unconditional OpenWebText generation.
• Score: 58.33696160279345
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Discrete diffusion models for categorical generation are defined by a corruption kernel, which determines the intermediate state space and the associated reverse prediction problem. We study uniform discrete diffusion and ask whether its training objective and reverse transitions can be enriched without changing the underlying categorical corruption process. We introduce Simplax, an exact Dirichlet--categorical augmentation that couples each corrupted categorical state with an auxiliary simplex-valued variable while preserving the original uniform diffusion process as its categorical marginal. This augmentation yields a tractable Rao--Blackwellized reverse-bridge objective and a corresponding stochastic reverse sampler, while retaining the corrupted categorical state as the denoiser input. Empirically, Simplax improves the generative perplexity--entropy tradeoff on unconditional OpenWebText generation. On Sudoku, a model trained exclusively on $30$-clue puzzles achieves the highest accuracy among the compared methods across all evaluated clue densities, including the minimum uniquely solvable $17$-clue regime, and also achieves the highest validity in unconditional generation.

Related papers

• Scale-Consistent Posterior Dynamics for Diffusion Inverse Problems [16.685352002333225]
  Posterior sampling with a pretrained diffusion prior is governed by a conditional score whose likelihood component is generally intractable.<n>We show that a noise-conditioned path whose targets approach the clean posterior produces a continuous surrogate SDE.<n>Experiments on FFHQ and ImageNet with 100 score evaluations demonstrate competitive reconstruction fidelity for superresolution and deblurring.
  arXiv  Detail & Related papers  (2026-08-15T09:38:41Z)
• What Does a Discrete Diffusion Model Learn? [71.03603607324338]
  discrete diffusion model learns: a denoiser, a score ratio, or a bridge plug-in predictor.<n>We start with a rigorous derivation of the continuous-time Markov chain (CTMC) ELBO for any noising process.<n>Every identity is verified numerically, without approximation, on an exactly solvable model.
  arXiv  Detail & Related papers  (2026-07-06T17:56:11Z)
• Efficient Sampling with Discrete Diffusion Models: Sharp and Adaptive Guarantees [9.180350432640912]
  We study the sampling efficiency of score-based discrete diffusion models under a continuous-time Markov chain (CTMC) formulation.<n>For uniform discrete diffusion, we show that the $$-leaping algorithm achieves an complexity of order $tilde O(d/varepsilon)$.<n>For masking discrete diffusion, we introduce a modified $$-leaping sampler whose convergence rate is governed by an intrinsic information-theoretic quantity.
  arXiv  Detail & Related papers  (2026-02-16T18:48:17Z)
• Tilt Matching for Scalable Sampling and Fine-Tuning [4.14348726233299]
  We propose a scalable algorithm for using interpolants to sample from unnormalized densities and for fine-tuning generative models.<n>The approach, Tilt Matching, arises from a dynamical equation relating the flow matching velocity to one targeting the same distribution tilted by a reward.<n>We empirically verify that the approach is efficient and highly scalable, providing state-of-the-art results on sampling under Lennard-Jones potentials and is competitive on fine-tuning Stable Diffusion.
  arXiv  Detail & Related papers  (2025-12-26T02:12:10Z)
• Effective Test-Time Scaling of Discrete Diffusion through Iterative Refinement [51.54933696252104]
  We introduce Iterative Reward-Guided Refinement (IterRef), a novel test-time scaling method tailored to discrete diffusion.<n>We formalize this process within a Multiple-Try Metropolis framework, proving convergence to the reward-aligned distribution.<n>IterRef achieves striking gains under low compute budgets, far surpassing prior state-of-the-art baselines.
  arXiv  Detail & Related papers  (2025-11-04T02:33:23Z)
• Self-diffusion for Solving Inverse Problems [3.8870795921263728]
  We propose self-diffusion, a novel framework for solving inverse problems without relying on pretrained generative models.<n>Self-diffusion exploits the spectral bias of neural networks and modulates it through a scheduled noise process.<n>We demonstrate the effectiveness of our approach on a variety of linear inverse problems, showing that self-diffusion achieves competitive or superior performance compared to other methods.
  arXiv  Detail & Related papers  (2025-10-24T12:57:22Z)
• Authentic Discrete Diffusion Model [72.31371542619121]
  Authentic Discrete Diffusion (ADD) framework redefines prior pseudo-discrete approaches.<n>ADD reformulates the diffusion input by directly using float-encoded one-hot class data.<n> experiments demonstrate that ADD achieves superior performance on classification tasks compared to the baseline.
  arXiv  Detail & Related papers  (2025-10-01T15:51:10Z)
• Rejection via Learning Density Ratios [50.91522897152437]
  Classification with rejection emerges as a learning paradigm which allows models to abstain from making predictions.<n>We propose a different distributional perspective, where we seek to find an idealized data distribution which maximizes a pretrained model's performance.<n>Our framework is tested empirically over clean and noisy datasets.
  arXiv  Detail & Related papers  (2024-05-29T01:32:17Z)
• DensePure: Understanding Diffusion Models towards Adversarial Robustness [110.84015494617528]
  We analyze the properties of diffusion models and establish the conditions under which they can enhance certified robustness. We propose a new method DensePure, designed to improve the certified robustness of a pretrained model (i.e. a classifier) We show that this robust region is a union of multiple convex sets, and is potentially much larger than the robust regions identified in previous works.
  arXiv  Detail & Related papers  (2022-11-01T08:18:07Z)
• False membership rate control in mixture models [1.387448620257867]
  A clustering task consists in partitioning elements of a sample into homogeneous groups. In the supervised setting, this approach is well known and referred to as classification with an abstention option. In this paper the approach is revisited in an unsupervised mixture model framework and the purpose is to develop a method that comes with the guarantee that the false membership rate does not exceed a pre-defined nominal level.
  arXiv  Detail & Related papers  (2022-03-04T22:37:59Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.