{"ID":2835824,"CreatedAt":"2026-06-01T04:54:23.091178241Z","UpdatedAt":"2026-06-01T04:54:23.091178241Z","DeletedAt":null,"paper_url":"https://arxiv.org/abs/2511.22177","arxiv_id":"2511.22177","title":"Designing Instance-Level Sampling Schedules via REINFORCE with James-Stein Shrinkage","abstract":"Most post-training methods for text-to-image samplers focus on model weights: either fine-tuning the backbone for alignment or distilling it for few-step efficiency. We take a different route: rescheduling the sampling timeline of a frozen sampler. Instead of a fixed, global schedule, we learn instance-level (prompt- and noise-conditioned) schedules through a single-pass Dirichlet policy. To ensure accurate gradient estimates in high-dimensional policy learning, we introduce a novel reward baseline based on a principled James-Stein estimator; it provably achieves lower estimation errors than commonly used variants and leads to superior performance. Our rescheduled samplers consistently improve text-image alignment including text rendering and compositional control across modern Stable Diffusion and Flux model families. Additionally, a 5-step Flux-Dev sampler with our schedules can attain generation quality comparable to deliberately distilled samplers like Flux-Schnell. We thus position our scheduling framework as an emerging model-agnostic post-training lever that unlocks additional generative potential in pretrained samplers.","short_abstract":"Most post-training methods for text-to-image samplers focus on model weights: either fine-tuning the backbone for alignment or distilling it for few-step efficiency. We take a different route: rescheduling the sampling timeline of a frozen sampler. Instead of a fixed, global schedule, we learn instance-level (prompt- a...","url_abs":"https://arxiv.org/abs/2511.22177","url_pdf":"https://arxiv.org/pdf/2511.22177v2","authors":"[\"Peiyu Yu\",\"Suraj Kothawade\",\"Sirui Xie\",\"Ying Nian Wu\",\"Hongliang Fei\"]","published":"2025-11-27T07:33:05Z","proceeding":"cs.LG","tasks":"[\"cs.LG\",\"cs.CV\"]","methods":"[\"Diffusion Model\"]","has_code":false}