English
Related papers

Related papers: Not All Preference Pairs Are Created Equal: A Reci…

200 papers

Aligning large language models (LLMs) with human values and intents critically involves the use of human or AI feedback. While dense feedback annotations are expensive to acquire and integrate, sparse feedback presents a structural design…

Machine Learning · Computer Science 2024-02-07 Hritik Bansal , John Dang , Aditya Grover

Preference learning, or the task of aligning generative models to preference comparison data, has yet to reach the conceptual maturity of classification, density estimation, etc. To close this gap, this work presents a framework to…

Machine Learning · Computer Science 2025-07-29 Lily H. Zhang , Rajesh Ranganath

Auto-annotation by ensemble of models is an efficient method of learning on unlabeled data. Wrong or inaccurate annotations generated by the ensemble may lead to performance degradation of the trained model. To deal with this problem we…

Computer Vision and Pattern Recognition · Computer Science 2024-03-14 Dror Simon , Miriam Farber , Roman Goldenberg

Recent work has demonstrated that pre-training in-domain language models can boost performance when adapting to a new domain. However, the costs associated with pre-training raise an important question: given a fixed budget, what steps…

Computation and Language · Computer Science 2022-05-16 Fan Bai , Alan Ritter , Wei Xu

Practitioners commonly align large language models using pairwise preferences, i.e., given labels of the type response A is preferred to response B for a given input. Perhaps less commonly, methods have also been developed for binary…

Computation and Language · Computer Science 2024-04-24 Jing Xu , Andrew Lee , Sainbayar Sukhbaatar , Jason Weston

Resource-efficient training optimization techniques are becoming increasingly important as the size of large language models (LLMs) continues to grow. In particular, batch packing is commonly used in pre-training and supervised fine-tuning…

Computation and Language · Computer Science 2026-03-02 Jaekyung Cho

High-quality data annotation is an essential but laborious and costly aspect of developing machine learning-based software. We explore the inherent tradeoff between annotation accuracy and cost by detecting and removing minority reports --…

Machine Learning · Computer Science 2025-04-15 Hsuan Wei Liao , Christopher Klugmann , Daniel Kondermann , Rafid Mahmood

Aligning large language models (LLMs) depends on high-quality datasets of human preference labels, which are costly to collect. Although active learning has been studied to improve sample efficiency relative to passive collection, many…

Machine Learning · Computer Science 2026-02-03 Yao Zhao , Kwang-Sung Jun

Preference optimization is widely used to align Large Language Models (LLMs) with preference feedback. However, most existing methods train on a single positive-negative pair per prompt, discarding additional supervision available in…

Computation and Language · Computer Science 2026-04-20 Jixuan Leng , Si Si , Hsiang-Fu Yu , Vinod Raman , Inderjit S. Dhillon

In-context learning is a promising paradigm that utilizes in-context examples as prompts for the predictions of large language models. These prompts are crucial for achieving strong performance. However, since the prompts need to be sampled…

Computation and Language · Computer Science 2025-07-15 Shaokun Zhang , Xiaobo Xia , Zhaoqing Wang , Ling-Hao Chen , Jiale Liu , Qingyun Wu , Tongliang Liu

Active learning aims to reduce annotation cost by predicting which samples are useful for a human expert to label. Although this field is quite old, several important challenges to using active learning in real-world settings still remain…

Machine Learning · Computer Science 2021-04-27 Louis Desreumaux , Vincent Lemaire

Direct Preference Optimization (DPO), a standard method for aligning language models with human preferences, is traditionally applied to offline preferences. Recent studies show that DPO benefits from iterative training with online…

Computation and Language · Computer Science 2024-06-18 Jie Liu , Zhanhui Zhou , Jiaheng Liu , Xingyuan Bu , Chao Yang , Han-Sen Zhong , Wanli Ouyang

Traditionally, most of the existing attribute learning methods are trained based on the consensus of annotations aggregated from a limited number of annotators. However, the consensus might fail in settings, especially when a wide spectrum…

Machine Learning · Computer Science 2019-06-19 Zhiyong Yang , Qianqian Xu , Xiaochun Cao , Qingming Huang

Offline preference optimization methods are efficient for large language models (LLMs) alignment. Direct Preference optimization (DPO)-like learning, one of the most popular approaches, stands out for its efficiency in reward modeling.…

Machine Learning · Computer Science 2026-05-26 Xiaobo Wang , Zixia Jia , Jiaqi Li , Qi Liu , Zilong Zheng

Linguistic bias in online news and social media is widespread but difficult to measure. Yet, its identification and quantification remain difficult due to subjectivity, context dependence, and the scarcity of high-quality gold-label…

Information Retrieval · Computer Science 2025-12-17 Fabian Haak , Philipp Schaer

Post-training processes are essential phases in grounding pre-trained language models to real-world tasks, with learning from demonstrations or preference signals playing a crucial role in this adaptation. We present a unified theoretical…

Machine Learning · Computer Science 2025-07-08 Bo Wang , Qinyuan Cheng , Runyu Peng , Rong Bao , Peiji Li , Qipeng Guo , Linyang Li , Zhiyuan Zeng , Yunhua Zhou , Xipeng Qiu

Direct preference optimization (DPO) is a successful fine-tuning strategy for aligning large language models with human preferences without the need to train a reward model or employ reinforcement learning. DPO, as originally formulated,…

Computation and Language · Computer Science 2024-06-07 Afra Amini , Tim Vieira , Ryan Cotterell

Aligning large language models (LLMs) is a central objective of post-training, often achieved through reward modeling and reinforcement learning methods. Among these, direct preference optimization (DPO) has emerged as a widely adopted…

Computation and Language · Computer Science 2026-03-03 Aladin Djuhera , Farhan Ahmed , Swanand Ravindra Kadhe , Syed Zawad , Heiko Ludwig , Holger Boche

Preference optimization methods such as DPO and KTO are widely used for aligning language models, yet little is understood about what properties of preference data drive downstream reasoning gains. We ask: what aspects of a preference pair…

Computation and Language · Computer Science 2026-04-13 Chia-Hsuan Lee , Mingyang Zhou , Renkun Ni , Zelei Cheng , Sihui Dai , Supriyo Chakraborty , Shixiong Zhang , Sambit Sahu , William Campbell

The development lifecycle of generative AI systems requires continual evaluation, data acquisition, and annotation, which is costly in both resources and time. In practice, rapid iteration often makes it necessary to rely on synthetic…

Machine Learning · Computer Science 2025-06-10 Anastasios N. Angelopoulos , Jacob Eisenstein , Jonathan Berant , Alekh Agarwal , Adam Fisch