English
Related papers

Related papers: Confronting Reward Overoptimization for Diffusion …

200 papers

Recent advancements in diffusion models (DMs) have been propelled by alignment methods that post-train models to better conform to human preferences. However, these approaches typically require computation-intensive training of a base model…

Computer Vision and Pattern Recognition · Computer Science 2025-08-05 Zejian Li , Yize Li , Chenye Meng , Zhongni Liu , Yang Ling , Shengyuan Zhang , Guang Yang , Changyuan Yang , Zhiyuan Yang , Lingyun Sun

Aligning large language models with human preferences has emerged as a critical focus in language modeling research. Yet, integrating preference learning into Text-to-Image (T2I) generative models is still relatively uncharted territory.…

Computer Vision and Pattern Recognition · Computer Science 2024-06-11 Yi Gu , Zhendong Wang , Yueqin Yin , Yujia Xie , Mingyuan Zhou

Direct preference optimization (DPO) has shown success in aligning diffusion models with human preference. Previous approaches typically assume a consistent preference label between final generations and noisy samples at intermediate steps,…

Machine Learning · Computer Science 2025-02-05 Jie Ren , Yuhang Zhang , Dongrui Liu , Xiaopeng Zhang , Qi Tian

Latent diffusion models are the state-of-the-art for synthetic image generation. To align these models with human preferences, training the models using reinforcement learning on human feedback is crucial. Black et. al 2024 introduced…

Machine Learning · Computer Science 2024-04-09 Mo Kordzanganeh , Danial Keshvary , Nariman Arian

Direct Preference Optimization (DPO) aligns text-to-image (T2I) generation models with human preferences using pairwise preference data. Although substantial resources are expended in collecting and labeling datasets, a critical aspect is…

Computer Vision and Pattern Recognition · Computer Science 2025-06-09 Yunhong Lu , Qichao Wang , Hengyuan Cao , Xiaoyin Xu , Min Zhang

Preference optimization for diffusion models aims to align them with human preferences for images. Previous methods typically use Vision-Language Models (VLMs) as pixel-level reward models to approximate human preferences. However, when…

Computer Vision and Pattern Recognition · Computer Science 2025-10-03 Tao Zhang , Cheng Da , Kun Ding , Huan Yang , Kun Jin , Yan Li , Tingting Gao , Di Zhang , Shiming Xiang , Chunhong Pan

Model-based reinforcement learning (RL) can be effectively supported at scale through the use of world models. However, in practice, scaling such approaches remains fundamentally limited. A commonly recognized challenge is model bias and…

Machine Learning · Computer Science 2026-05-27 Xiaoyuan Cheng , Wenxuan Yuan , Zhancun Mu , Yuanzhao Zhang , Yiming Yang , Hai Wang , Zhuo Sun , Che Liu

This tutorial provides an in-depth guide on inference-time guidance and alignment methods for optimizing downstream reward functions in diffusion models. While diffusion models are renowned for their generative modeling capabilities,…

Artificial Intelligence · Computer Science 2025-01-22 Masatoshi Uehara , Yulai Zhao , Chenyu Wang , Xiner Li , Aviv Regev , Sergey Levine , Tommaso Biancalani

Language model (LM) post-training (or alignment) involves maximizing a reward function that is derived from preference annotations. Direct Preference Optimization (DPO) is a popular offline alignment method that trains a policy directly on…

Machine Learning · Computer Science 2025-03-04 Adam Fisch , Jacob Eisenstein , Vicky Zayats , Alekh Agarwal , Ahmad Beirami , Chirag Nagpal , Pete Shaw , Jonathan Berant

Diffusion models have shown promising generative capabilities across diverse domains, yet aligning their outputs with desired reward functions remains a challenge, particularly in cases where reward functions are non-differentiable. Some…

Machine Learning · Computer Science 2025-06-04 Xiner Li , Masatoshi Uehara , Xingyu Su , Gabriele Scalia , Tommaso Biancalani , Aviv Regev , Sergey Levine , Shuiwang Ji

Few-step diffusion models enable efficient high-resolution image synthesis but struggle to align with specific downstream objectives due to limitations of existing reinforcement learning (RL) methods in low-step regimes with limited state…

Machine Learning · Computer Science 2026-03-02 Ziyi Zhang , Li Shen , Sen Zhang , Deheng Ye , Yong Luo , Miaojing Shi , Dongjing Shan , Bo Du , Dacheng Tao

Reinforcement Learning from Human Feedback has emerged as a standard for aligning diffusion models. However, we identify a fundamental limitation in the standard DPO formulation because it relies on the Bradley-Terry model to aggregate…

Computer Vision and Pattern Recognition · Computer Science 2025-12-12 Jiho Jang , Jinyoung Kim , Kyungjune Baek , Nojun Kwak

To fully leverage the capabilities of diffusion models, we are often interested in optimizing downstream reward functions during inference. While numerous algorithms for reward-guided generation have been recently proposed due to their…

Machine Learning · Computer Science 2025-04-18 Masatoshi Uehara , Xingyu Su , Yulai Zhao , Xiner Li , Aviv Regev , Shuiwang Ji , Sergey Levine , Tommaso Biancalani

Diffusion models have achieved remarkable success in generating realistic and versatile images from text prompts. Inspired by the recent advancements of language models, there is an increasing interest in further improving the models by…

Computer Vision and Pattern Recognition · Computer Science 2025-10-07 Binxu Li , Minkai Xu , Jiaqi Han , Meihua Dang , Stefano Ermon

Online reinforcement learning is becoming increasingly important for aligning diffusion models with non-differentiable objectives. However, existing methods still face limitations in assigning fine-grained credit along denoising…

Machine Learning · Computer Science 2026-05-28 Zhengyang Liang , Qihang Zhang , Ceyuan Yang

We introduce Adversarial Policy Optimization (AdvPO), a novel solution to the pervasive issue of reward over-optimization in Reinforcement Learning from Human Feedback (RLHF) for Large Language Models (LLMs). Over-optimization occurs when a…

Machine Learning · Computer Science 2024-07-10 Xiaoying Zhang , Jean-Francois Ton , Wei Shen , Hongning Wang , Yang Liu

In offline reinforcement learning, value overestimation caused by out-of-distribution (OOD) actions significantly limits policy performance. Recently, diffusion models have been leveraged for their strong distribution-matching capabilities,…

Machine Learning · Computer Science 2025-11-13 Yunchang Ma , Tenglong Liu , Yixing Lan , Xin Yin , Changxin Zhang , Xinglong Zhang , Xin Xu

Direct Preference Optimization (DPO) has been successfully used to align large language models (LLMs) according to human preferences, and more recently it has also been applied to improving the quality of text-to-image diffusion models.…

Computer Vision and Pattern Recognition · Computer Science 2025-10-07 Shivanshu Shekhar , Shreyas Singh , Tong Zhang

Diffusion models have achieved remarkable success in text-to-image generation. However, their practical applications are hindered by the misalignment between generated images and corresponding text prompts. To tackle this issue,…

Computer Vision and Pattern Recognition · Computer Science 2025-03-28 Zijing Hu , Fengda Zhang , Long Chen , Kun Kuang , Jiahui Li , Kaifeng Gao , Jun Xiao , Xin Wang , Wenwu Zhu

The new paradigm of test-time scaling has yielded remarkable breakthroughs in Large Language Models (LLMs) (e.g. reasoning models) and in generative vision models, allowing models to allocate additional computation during inference to…

Machine Learning · Computer Science 2025-08-14 Luca Eyring , Shyamgopal Karthik , Alexey Dosovitskiy , Nataniel Ruiz , Zeynep Akata