从离散扩散模型中的全局专家组合到因子级专家组合

中文摘要 AI 辅助

离散扩散模型为解决复杂推理任务提供了强大框架，特别是通过组合生成，将多个预训练专家结合以超越其各自训练数据进行泛化。近期理论修正引入了随时间变化的混合权重，以更好地使组合扩散动力学与预期目标对齐。然而，这些方法基于逐个样本工作，整体处理每个生成状态，忽略了不同专家潜在的空间或功能专业化。在本文中，我们提出了FactorDiff——一种用于扩散模型的因子级组合框架，以解决此限制。我们认为样本可进一步分解为更小因子，并提出一种采样过程，将每个因子动态路由到最相关专家。我们用空间/像素级组合实例化此框架，并在ARC - AGI基准上进行验证，表明在需要逻辑一致性和空间解缠的任务上，简单的因子特定路由始终优于复杂的全局标量加权方案。

英文摘要

Discrete diffusion models offer a powerful framework for solving complex reasoning tasks, particularly through compositional generation, which combines multiple pre-trained experts to generalize beyond their individual training data. Recent theoretical corrections introduce time-dependent mixing weights to better align composed diffusion dynamics with the intended target. However, these methods are fundamentally limited by working on a per-sample basis, treating each generated state monolithically and ignoring the potential spatial or functional specializations of different experts. In this work, we address this limitation by proposing FactorDiff - a factor-wise composition framework for diffusion models. We posit that samples can be further decomposed into smaller factors, and propose a sampling process that dynamically routes each factor to the most relevant expert. We instantiate this framework with spatial/pixel-level compositions and validate it on the ARC-AGI benchmark, demonstrating that simple factor-specific routing consistently outperforms complex global scalar weighting schemes on tasks that require logical consistency and spatial disentanglement.

发表机构

• University of Toronto（多伦多大学）
• Vector Institute for Artificial Intelligence（向量人工智能研究所）
• Mechanical & Industrial Engineering, University of Toronto（多伦多大学机械与工业工程系）
• Canadian Institute for Advanced Research (CIFAR)（加拿大高级研究所）

机构由 AI 辅助整理，请以论文原文为准。