English
Related papers

Related papers: SparsePO: Controlling Preference Alignment of LLMs…

200 papers

Large Language Models (LLMs) have demonstrated unprecedented generative capabilities, yet their alignment with human values remains critical for ensuring helpful and harmless deployments. While Reinforcement Learning from Human Feedback…

Direct Preference Optimization is an offline post-SFT method for aligning language models from preference pairs, with strong results in instruction following and summarization. However, DPO's sequence-level implicit reward can be brittle…

Computation and Language · Computer Science 2026-03-03 Samah Fodeh , Linhai Ma , Ganesh Puthiaraju , Srivani Talakokkul , Afshan Khan , Ashley Hagaman , Sarah R. Lowe , Aimee Kendall Roundtree

In the post-training of large language models (LLMs), Reinforcement Learning from Human Feedback (RLHF) is an effective approach to achieve generation aligned with human preferences. Direct Preference Optimization (DPO) allows for policy…

Machine Learning · Computer Science 2025-06-16 Motoki Omura , Yasuhiro Fujita , Toshiki Kataoka

The recent success in using human preferences to align large language models (LLMs) has significantly improved their performance in various downstream tasks, such as question answering, mathematical reasoning, and code generation. However,…

Machine Learning · Computer Science 2026-05-18 Xiaoqiang Lin , Arun Verma , Zhongxiang Dai , Daniela Rus , See-Kiong Ng , Bryan Kian Hsiang Low

Optimizing policies based on human preferences is key to aligning language models with human intent. This work focuses on reward modeling, a core component in reinforcement learning from human feedback (RLHF), and offline preference…

Machine Learning · Computer Science 2025-06-02 Soichiro Nishimori , Yu-Jie Zhang , Thanawat Lodkaew , Masashi Sugiyama

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

Direct Preference Optimization (DPO) demonstrates the advantage of aligning a large language model with human preference using only an offline dataset. However, DPO has the limitation that the KL penalty, which prevents excessive deviation…

Machine Learning · Computer Science 2025-10-28 Sangkyu Lee , Janghoon Han , Hosung Song , Stanley Jungkyu Choi , Honglak Lee , Youngjae Yu

The next token prediction loss is the dominant self-supervised training objective for large language models and has achieved promising results in a variety of downstream tasks. However, upon closer investigation of this objective, we find…

Computation and Language · Computer Science 2025-02-25 Zhili Feng , Dhananjay Ram , Cole Hawkins , Aditya Rawal , Jinman Zhao , Sheng Zha

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

Direct Preference Optimization (DPO) has emerged as a predominant alignment method for diffusion models, facilitating off-policy training without explicit reward modeling. However, its reliance on large-scale, high-quality human preference…

Computer Vision and Pattern Recognition · Computer Science 2026-02-09 Khiem Pham , Quang Nguyen , Tung Nguyen , Jingsen Zhu , Michele Santacatterina , Dimitris Metaxas , Ramin Zabih

Aligning large language models (LLMs) with diverse and multifaceted user preferences is a fundamental challenge in personalized AI systems. Existing multi-objective alignment methods either rely on costly training or require pre-trained…

Computation and Language · Computer Science 2026-05-26 Linhao Luo , Thuy-Trang Vu , Van-Anh Nguyen , Junae Kim , Gholamreza Haffari , Dinh Phung

A single language model, even when aligned with labelers through reinforcement learning from human feedback (RLHF), may not suit all human preferences. Recent approaches therefore prefer customization, gathering multi-dimensional feedback,…

Machine Learning · Computer Science 2024-08-20 Zhanhui Zhou , Jie Liu , Jing Shao , Xiangyu Yue , Chao Yang , Wanli Ouyang , Yu Qiao

Most Video Large Language Models (Video-LLMs) adopt preference alignment techniques, e.g., DPO~\citep{rafailov2024dpo}, to optimize the reward margin between a winning response ($y_w$) and a losing response ($y_l$). However, the likelihood…

Computer Vision and Pattern Recognition · Computer Science 2025-06-06 Xiaodong Wang , Jinfa Huang , Li Yuan , Peixi Peng

A common technique for aligning large language models (LLMs) relies on acquiring human preferences by comparing multiple generations conditioned on a fixed context. This method, however, relies solely on pairwise comparisons, where the…

Computation and Language · Computer Science 2025-01-09 Hritik Bansal , Ashima Suvarna , Gantavya Bhatt , Nanyun Peng , Kai-Wei Chang , Aditya Grover

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

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

Direct Preference Optimization (DPO) is broadly utilized for aligning Large Language Models (LLMs) with human values because of its flexibility. Despite its effectiveness, it has been observed that the capability of DPO to generate…

Machine Learning · Computer Science 2025-05-20 Wenqiao Zhu , Ji Liu , Lulu Wang , Jun Wu , Yulun Zhang

In the domain of complex reasoning tasks, such as mathematical reasoning, recent advancements have proposed the use of Direct Preference Optimization (DPO) to suppress output of dispreferred responses, thereby enhancing the long-chain…

Computation and Language · Computer Science 2025-10-27 Weibin Liao , Xu Chu , Yasha Wang

Direct Preference Optimization (DPO) is widely utilized in the Reinforcement Learning from Human Feedback (RLHF) phase to align Large Language Models (LLMs) with human preferences, thereby enhancing both their harmlessness and efficacy.…

Machine Learning · Computer Science 2024-12-02 Wei Liu , Yang Bai , Chengcheng Han , Rongxiang Weng , Jun Xu , Xuezhi Cao , Jingang Wang , Xunliang Cai

The increasing capabilities of large language models (LLMs) raise opportunities for artificial general intelligence but concurrently amplify safety concerns, such as potential misuse of AI systems, necessitating effective AI alignment.…

Machine Learning · Computer Science 2023-09-29 Chaoqi Wang , Yibo Jiang , Chenghao Yang , Han Liu , Yuxin Chen