{"slug": "d2po-optimizing-diffusion-samplers-via-dynamic-preference", "title": "D2PO: Optimizing Diffusion Samplers via Dynamic Preference", "summary": "Researchers propose D2PO, a framework that optimizes diffusion sampling policies by reformulating sampler optimization as a preference-based alignment problem using Direct Preference Optimization (DPO). D2PO introduces dynamic preferences and an energy-based model to improve perceptual quality, outperforming conventional regression-based schedulers under low-NFE constraints.", "body_md": "arXiv:2607.06609v1 Announce Type: new\nAbstract: We propose D2PO (Dynamic Direct Preference Optimization), a principled framework for optimizing diffusion sampling policies with respect to timestep schedules and classifier-free guidance (CFG) weights. Our work is motivated by a fundamental limitation of existing student-teacher regression frameworks; low-NFE student samplers are trained to mimic high-NFEteachers, often sacrificing high-frequency texture fidelity while preserving coarse global structures, thereby misaligning the sampler with perceptual quality. D2PO addresses this challenge by reformulating sampler optimization as a preference-based alignment problem, leveraging the Direct Preference Optimization (DPO) framework. To make DPO applicable to diffusion samplers, we model the sampling policy as an energy-based model (EBM), transforming preference comparisons into tractable energy differences. We further introduce a novel energy formulation derived directly from the pretrained score network, enabling preference evaluation in perturbed spaces that jointly capture structural consistency and fine-grained details. Moreover, we introduce dynamic preferences, where the preferred samples used for alignment progressively improve as the sampling policies are learned. This self-improving mechanism replaces rigid static teacher supervision with an iterative, preference-guided refinement process, providing progressively stronger alignment signals. Extensive experiments demonstrate that D2PO aligns diffusion samplers with perceptual quality more faithfully, unlocking the full potential of high-quality teachers and consistently outperforming conventional regression-based schedulers under low-NFE constraints.", "url": "https://wpnews.pro/news/d2po-optimizing-diffusion-samplers-via-dynamic-preference", "canonical_source": "https://arxiv.org/abs/2607.06609", "published_at": "2026-07-09 04:00:00+00:00", "updated_at": "2026-07-09 04:24:49.558054+00:00", "lang": "en", "topics": ["artificial-intelligence", "machine-learning", "generative-ai", "ai-research"], "entities": ["D2PO", "Direct Preference Optimization", "arXiv"], "alternates": {"html": "https://wpnews.pro/news/d2po-optimizing-diffusion-samplers-via-dynamic-preference", "markdown": "https://wpnews.pro/news/d2po-optimizing-diffusion-samplers-via-dynamic-preference.md", "text": "https://wpnews.pro/news/d2po-optimizing-diffusion-samplers-via-dynamic-preference.txt", "jsonld": "https://wpnews.pro/news/d2po-optimizing-diffusion-samplers-via-dynamic-preference.jsonld"}}