paper-with-me

홈 › Papers

Directly Fine-Tuning Diffusion Models on Differentiable Rewards

2023-09-29 · Kevin Clark, Paul Vicol, Kevin Swersky, David J Fleet

We present Direct Reward Fine-Tuning (DRaFT), a simple and effective method for fine-tuning diffusion models to maximize differentiable reward functions, such as scores from human preference models. We first show that it is possible to backpropagate the reward function gradient through the full sampling procedure, and that doing so achieves strong performance on a variety of rewards, outperforming reinforcement learning-based approaches. We then propose more efficient variants of DRaFT: DRaFT-K, which truncates backpropagation to only the last K steps of sampling, and DRaFT-LV, which obtains lower-variance gradient estimates for the case when K=1. We show that our methods work well for a variety of reward functions and can be used to substantially improve the aesthetic quality of images generated by Stable Diffusion 1.4. Finally, we draw connections between our approach and prior work, providing a unifying perspective on the design space of gradient-based fine-tuning algorithms.

📄 PDF Abstract BibTeX arXiv:2309.17400

Code (1)

nvidia/nemo-aligner pytorch

Methods 이 논문이 사용한 방법론

Diffusion Diffusion models generate samples by gradually removing noise from a signal, and their training objective can be expressed as a reweighted variational lower-bound…

Similar Papers 제목 키워드 기반

Diffusion-DRF: Free, Rich, and Differentiable Reward for Video Diffusion Fine-Tuning

2026-01-07 · Yifan Wang, Yanyu Li, Gordon Guocheng Qian, Sergey Tulyakov 외 arxiv

Video diffusion alignment has been heavily relied on scalar rewards. These rewards are typically derived from learned reward models in human preference datasets, requiring additional training and extensive collection. Mo…

Reward Fine-Tuning Two-Step Diffusion Models via Learning Differentiable Latent-Space Surrogate Reward

2024-11-22 · CVPR 2025 1 · Zhiwei Jia, Yuesong Nan, Huixi Zhao, Gengdai Liu

Recent research has shown that fine-tuning diffusion models (DMs) with arbitrary rewards, including non-differentiable ones, is feasible with reinforcement learning (RL) techniques, enabling flexible model alignment. How…

Image GenerationReinforcement Learning (RL)

Directly Aligning the Full Diffusion Trajectory with Fine-Grained Human Preference

2025-09-08 · Xiangwei Shen, Zhimin Li, Zhantao Yang, Shiyi Zhang 외 arxiv

Recent studies have demonstrated the effectiveness of directly aligning diffusion models with human preferences using differentiable reward. However, they exhibit two primary challenges: (1) they rely on multistep denois…

Derivative-Free Guidance in Continuous and Discrete Diffusion Models with Soft Value-Based Decoding

2024-08-15 · Xiner Li, Yulai Zhao, Chenyu Wang, Gabriele Scalia 외

Diffusion models excel at capturing the natural design spaces of images, molecules, DNA, RNA, and protein sequences. However, rather than merely generating designs that are natural, we often aim to optimize downstream re…

Image Generation

VARD: Efficient and Dense Fine-Tuning for Diffusion Models with Value-based RL

2025-05-21 · Fengyuan Dai, Zifeng Zhuang, Yufei Huang, Siteng Huang 외

Diffusion models have emerged as powerful generative tools across various domains, yet tailoring pre-trained models to exhibit specific desirable properties remains challenging. While reinforcement learning (RL) offers a…

Reinforcement Learning (RL)