paper-with-me

홈 › Papers

Improved techniques for fine-tuning flow models via adjoint matching: a deterministic control pipeline

2026-05-07 · Zhengyi Guo, Jiayuan Sheng, David D. Yao, Wenpin Tang arxiv

We propose a deterministic adjoint matching framework that formulates human preference alignment for flow-based generative models as an optimal control problem over velocity fields. One can directly regress the control toward a value-gradient-induced target under the current policy, leading to a simple and stable training objective. Building on this perspective, we introduce a truncated adjoint scheme that focuses computation on the terminal portion of the trajectory, where reward-relevant signals concentrate, which yields substantial computational savings while preserving alignment quality. We further generalize the framework beyond standard KL-based regularization, allowing more flexible trade-offs between alignment strength and distributional preservation. Experiments on SiT-XL/2 and FLUX.2-Klein-4B demonstrate consistent gains across multiple alignment metrics, along with substantially improved diversity and mode preservation.

📄 PDF Abstract BibTeX arXiv:2605.06583

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

A unified perspective on fine-tuning and sampling with diffusion and flow models

2026-04-30 · Carles Domingo-Enrich, Yuanqi Du, Michael S. Albergo arxiv

We study the problem of training diffusion and flow generative models to sample from target distributions defined by an exponential tilting of a base density; a formulation that subsumes both sampling from unnormalized d…

Efficient Adjoint Matching for Fine-tuning Diffusion Models

2026-05-12 · Jeongwoo Shin, Dongsoo Shin, Yuchen Zhu, Wei Guo 외 arxiv

Reward fine-tuning has become a common approach for aligning pretrained diffusion and flow models with human preferences in text-to-image generation. Among reward-gradient-based methods, Adjoint Matching (AM) provides a …

Text-to-Image Generation

Adjoint Matching: Fine-tuning Flow and Diffusion Generative Models with Memoryless Stochastic Optimal Control

2024-09-13 · Carles Domingo-Enrich, Michal Drozdzal, Brian Karrer, Ricky T. Q. Chen

Dynamical generative models that produce samples through an iterative process, such as Flow Matching and denoising diffusion models, have seen widespread use, but there have not been many theoretically-sound methods for …

DenoisingDiversity

AdjointDPM: Adjoint Sensitivity Method for Gradient Backpropagation of Diffusion Probabilistic Models

2023-07-20 · Jiachun Pan, Jun Hao Liew, Vincent Y. F. Tan, Jiashi Feng 외

Existing customization methods require access to multiple reference examples to align pre-trained diffusion probabilistic models (DPMs) with user-provided concepts. This paper aims to address the challenge of DPM customi…

DenoisingSensitivity

Flows and Diffusions on the Neural Manifold

2025-07-14 · Daniel Saragih, Deyu Cao, Tejas Balaji arxiv

Diffusion and flow-based generative models have achieved remarkable success in domains such as image synthesis, video generation, and natural language modeling. In this work, we extend these advances to weight space lear…

Video Generation