Debiasing Guidance for Discrete Diffusion with Sequential Monte Carlo

• URL: http://arxiv.org/abs/2502.06079v2
• Date: Mon, 17 Feb 2025 15:44:24 GMT
• Title: Debiasing Guidance for Discrete Diffusion with Sequential Monte Carlo
• Authors: Cheuk Kit Lee, Paul Jeha, Jes Frellsen, Pietro Lio, Michael Samuel Albergo, Francisco Vargas,
• Abstract summary: We introduce a Sequential Monte Carlo algorithm that generates unbiasedly from a target distribution. We validate our approach on low-dimensional distributions, controlled images and text generations.
• Score: 10.948453531321032
• License:
• Abstract: Discrete diffusion models are a class of generative models that produce samples from an approximated data distribution within a discrete state space. Often, there is a need to target specific regions of the data distribution. Current guidance methods aim to sample from a distribution with mass proportional to $p_0(x_0) p(\zeta|x_0)^\alpha$ but fail to achieve this in practice. We introduce a Sequential Monte Carlo algorithm that generates unbiasedly from this target distribution, utilising the learnt unconditional and guided process. We validate our approach on low-dimensional distributions, controlled images and text generations. For text generation, our method provides strong control while maintaining low perplexity compared to guidance-based approaches.

Related papers

• Reverse Markov Learning: Multi-Step Generative Models for Complex Distributions [10.165179181394755]
  We extend engression to improve its capability in learning complex distributions. We propose a framework that defines a general forward process transitioning from the target distribution to a known distribution. This reverse process reconstructs the target distribution step by step.
  arXiv  Detail & Related papers  (2025-02-19T14:10:15Z)
• Designing a Conditional Prior Distribution for Flow-Based Generative Models [16.729797131896138]
  Flow-basedgenerative models have recently shown impressive performance for conditional generation tasks. In this work, we tap into a non-utilized property of conditional flow-based models: the ability to design a non-trivial prior distribution. We utilize the flow matching formulation to map samples from a parametric distribution centered around this point to the conditional target distribution.
  arXiv  Detail & Related papers  (2025-02-13T18:58:15Z)
• Distributional Diffusion Models with Scoring Rules [83.38210785728994]
  Diffusion models generate high-quality synthetic data. generating high-quality outputs requires many discretization steps. We propose to accomplish sample generation by learning the posterior em distribution of clean data samples.
  arXiv  Detail & Related papers  (2025-02-04T16:59:03Z)
• Theory on Score-Mismatched Diffusion Models and Zero-Shot Conditional Samplers [49.97755400231656]
  We present the first performance guarantee with explicit dimensional general score-mismatched diffusion samplers. We show that score mismatches result in an distributional bias between the target and sampling distributions, proportional to the accumulated mismatch between the target and training distributions. This result can be directly applied to zero-shot conditional samplers for any conditional model, irrespective of measurement noise.
  arXiv  Detail & Related papers  (2024-10-17T16:42:12Z)
• Conditional sampling within generative diffusion models [12.608803080528142]
  We present a review of existing computational approaches to conditional sampling within generative diffusion models. We highlight key methodologies that either utilise the joint distribution, or rely on (pre-trained) marginal distributions with explicit likelihoods.
  arXiv  Detail & Related papers  (2024-09-15T07:48:40Z)
• Unsupervised Learning of Sampling Distributions for Particle Filters [80.6716888175925]
  We put forward four methods for learning sampling distributions from observed measurements. Experiments demonstrate that learned sampling distributions exhibit better performance than designed, minimum-degeneracy sampling distributions.
  arXiv  Detail & Related papers  (2023-02-02T15:50:21Z)
• SIXO: Smoothing Inference with Twisted Objectives [8.049531918823758]
  We introduce SIXO, a method that learns targets that approximate the smoothing distributions. We then use SMC with these learned targets to define a variational objective for model and proposal learning.
  arXiv  Detail & Related papers  (2022-06-13T07:46:35Z)
• Sampling from Arbitrary Functions via PSD Models [55.41644538483948]
  We take a two-step approach by first modeling the probability distribution and then sampling from that model. We show that these models can approximate a large class of densities concisely using few evaluations, and present a simple algorithm to effectively sample from these models.
  arXiv  Detail & Related papers  (2021-10-20T12:25:22Z)
• Distributional Reinforcement Learning via Moment Matching [54.16108052278444]
  We formulate a method that learns a finite set of statistics from each return distribution via neural networks. Our method can be interpreted as implicitly matching all orders of moments between a return distribution and its Bellman target. Experiments on the suite of Atari games show that our method outperforms the standard distributional RL baselines.
  arXiv  Detail & Related papers  (2020-07-24T05:18:17Z)
• Stein Variational Inference for Discrete Distributions [70.19352762933259]
  We propose a simple yet general framework that transforms discrete distributions to equivalent piecewise continuous distributions. Our method outperforms traditional algorithms such as Gibbs sampling and discontinuous Hamiltonian Monte Carlo. We demonstrate that our method provides a promising tool for learning ensembles of binarized neural network (BNN) In addition, such transform can be straightforwardly employed in gradient-free kernelized Stein discrepancy to perform goodness-of-fit (GOF) test on discrete distributions.
  arXiv  Detail & Related papers  (2020-03-01T22:45:41Z)

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.