English
Related papers

Related papers: Efficient Preference Poisoning Attack on Offline R…

200 papers

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

Reinforcement Learning from Human Feedback (RLHF) and derivative techniques like Direct Preference Optimization (DPO) are task-alignment algorithms used to repurpose general, foundational models for specific tasks. We show that applying…

Computation and Language · Computer Science 2025-09-30 Kaden Uhlig , Joern Wuebker , Raphael Reinauer , John DeNero

In a backdoor attack, an adversary injects corrupted data into a model's training dataset in order to gain control over its predictions on images with a specific attacker-defined trigger. A typical corrupted training example requires…

Machine Learning · Computer Science 2023-10-31 Rishi D. Jha , Jonathan Hayase , Sewoong Oh

Reinforcement Learning from Human Feedback (RLHF) is crucial for aligning text-to-image (T2I) models with human preferences. However, RLHF's feedback mechanism also opens new pathways for adversaries. This paper demonstrates the feasibility…

Machine Learning · Computer Science 2025-06-05 Kaiwen Duan , Hongwei Yao , Yufei Chen , Ziyun Li , Tong Qiao , Zhan Qin , Cong Wang

Label manipulation attacks are a subclass of data poisoning attacks in adversarial machine learning used against different applications, such as malware detection. These types of attacks represent a serious threat to detection systems in…

Machine Learning · Computer Science 2020-06-17 Rahim Taheri , Reza Javidan , Mohammad Shojafar , Zahra Pooranian , Ali Miri , Mauro Conti

Preference optimization is a critical post-training technique used to align large language models (LLMs) with human preferences, typically by fine-tuning on ranked response pairs. While methods like Direct Preference Optimization (DPO) have…

Computation and Language · Computer Science 2025-11-12 Rhitabrat Pokharel , Yufei Tao , Ameeta Agrawal

With the increase in machine learning (ML) applications in different domains, incentives for deceiving these models have reached more than ever. As data is the core backbone of ML algorithms, attackers shifted their interest toward…

Cryptography and Security · Computer Science 2023-01-04 Kshitiz Aryal , Maanak Gupta , Mahmoud Abdelsalam

Reinforcement learning with human feedback~(RLHF) is critical for aligning Large Language Models (LLMs) with human preference. Compared to the widely studied offline version of RLHF, \emph{e.g.} direct preference optimization (DPO), recent…

Machine Learning · Computer Science 2024-06-13 Lichang Chen , Jiuhai Chen , Chenxi Liu , John Kirchenbauer , Davit Soselia , Chen Zhu , Tom Goldstein , Tianyi Zhou , Heng Huang

Reinforcement learning with human feedback (RLHF) has become the dominant method to align large models to user preferences. Unlike fine-tuning, for which there are many studies regarding training data memorization, it is not clear how…

Machine Learning · Computer Science 2024-10-28 Aneesh Pappu , Billy Porter , Ilia Shumailov , Jamie Hayes

Direct Preference Optimization (DPO) has emerged as a cornerstone of reinforcement learning from human feedback (RLHF) due to its simplicity and efficiency. However, existing DPO-based methods typically treat all preference pairs equally,…

Machine Learning · Computer Science 2026-05-26 Shangpin Peng , Weinong Wang , Zhuotao Tian , Senqiao Yang , Xing Wu , Haotian Xu , Chengquan Zhang , Takashi Isobe , Baotian Hu , Min Zhang

Direct Preference Optimization (DPO) guides large language models (LLMs) to generate recommendations aligned with user historical behavior distributions by minimizing preference alignment loss. However, our systematic empirical research and…

Information Retrieval · Computer Science 2026-05-28 Chu Zhao , Enneng Yang , Jianzhe Zhao , Guibing Guo

Direct Preference Optimization (DPO), the efficient alternative to PPO-based RLHF, falls short on knowledge-intensive generation: standard preference signals from human annotators or LLM judges exhibit a systematic verbosity bias that…

Computation and Language · Computer Science 2026-05-13 Qiming Bao , Juho Leinonen , Paul Denny , Michael J. Witbrock

Direct Preference Optimization (DPO) have emerged as a popular method for aligning Large Language Models (LLMs) with human preferences. While DPO effectively preserves the relative ordering between chosen and rejected responses through…

Computation and Language · Computer Science 2025-06-05 Lin Sun , Chuang Liu , Peng Liu , Bingyang Li , Weijia Lu , Ning Wu

To explore the vulnerability of deep neural networks (DNNs), many attack paradigms have been well studied, such as the poisoning-based backdoor attack in the training stage and the adversarial attack in the inference stage. In this paper,…

Cryptography and Security · Computer Science 2022-07-27 Jiawang Bai , Baoyuan Wu , Zhifeng Li , Shu-tao Xia

Reinforcement learning from human feedback (RLHF) provides a principled framework for aligning AI systems with human preference data. For various reasons, e.g., personal bias, context ambiguity, lack of training, etc, human annotators may…

Machine Learning · Computer Science 2024-07-10 Alexander Bukharin , Ilgee Hong , Haoming Jiang , Zichong Li , Qingru Zhang , Zixuan Zhang , Tuo Zhao

Direct alignment algorithms have proven an effective step for aligning language models to human-desired behaviors. Current variants of the Direct Preference Optimization objective have focused on a strict setting where all tokens are…

Computation and Language · Computer Science 2025-11-03 Fenia Christopoulou , Ronald Cardenas , Gerasimos Lampouras , Haitham Bou-Ammar , Jun Wang

Despite the efficacy of Direct Preference Optimization (DPO) in aligning Large Language Models (LLMs), reward hacking remains a pivotal challenge. This issue emerges when LLMs excessively reduce the probability of rejected completions to…

Computation and Language · Computer Science 2025-08-26 Chenxu Yang , Ruipeng Jia , Mingyu Zheng , Naibin Gu , Zheng Lin , Siyuan Chen , Weichong Yin , Hua Wu , Weiping Wang

Offline preference optimization offers a simpler and more stable alternative to RLHF for aligning language models. However, their effectiveness is critically dependent on ranking accuracy, a metric where further gains are highly impactful.…

Computation and Language · Computer Science 2025-11-18 Ruibo Deng , Duanyu Feng , Wenqiang Lei

Efficient preference optimization algorithms such as Direct Preference Optimization (DPO) have become a popular approach in aligning large language models (LLMs) with human preferences. These algorithms implicitly treat the LLM as a reward…

Computation and Language · Computer Science 2025-07-29 Tong Liu , Xiao Yu , Wenxuan Zhou , Jindong Gu , Volker Tresp

Reinforcement learning from human feedback (RLHF) has been popular for aligning text-to-image (T2I) diffusion models with human preferences. As a mainstream branch of RLHF, Direct Preference Optimization (DPO) offers a computationally…

Machine Learning · Computer Science 2026-05-07 Jiaming Hu , Jiamu Bai , Haoyu Wang , Debarghya Mukherjee , Ioannis Ch. Paschalidis