The paper introduces Cross‑lingual Ranking Preference Optimization (CRPO), a framework that uses high‑quality English preference data to improve alignment of large language models in other languages. CRPO builds a hierarchical structure over parallel preference pairs, jointly optimizing intra‑ and inter‑lingual preferences and providing a relative ranking signal beyond binary comparisons. Experiments on five languages show consistent gains in instruction‑following and knowledge utilization, with robust performance across different weighting schemes and improved reward margins and log‑probabilities of desirable responses.
By Seungyoon Lee, Minhyuk Kim, Jungseob Lee, Heuiseok Lim
CroCo introduces cross‑lingual contrastive preference tuning on self‑generations, extending prior English‑only methods to 14 high‑ and low‑resource languages. A reward model trained solely on English preferences, applied to a multilingual base, yields effective within‑language rankings and improves performance in both monolingual and multilingual settings without catastrophic forgetting. The approach requires on‑policy data; off‑policy responses and online preference optimization offer limited gains, yet on structured tasks CroCo matches or surpasses the base model in most languages, and on open‑ended generation it wins 28/30 judge evaluations across 15 languages.
By Mike Zhang, Ali Basirat, Desmond Elliott
Aligning large language models to human preferences is crucial for real-world deployment but frequently incurs an alignment tax, leading to the catastrophic forgetting of pre-trained general capabilit...
arXiv:2602. 02898v3 Announce Type: replace Abstract: Language model benchmarks are pervasive and computationally-efficient proxies for real-world performance.
By Marco Gutierrez, Xinyi Leng, Hannah Cyberey, Jonathan Richard Schwarz, Ahmed Alaa, Thomas Hartvigsen
The paper introduces BALIGN, a balanced data selection strategy designed to reduce catastrophic forgetting—referred to as the alignment tax—in large language models during preference-based alignment. By analyzing preference optimization gradients, the authors identify three data-centric features that influence parameter drift: the reference model's log-probability margin, token length differences between chosen and rejected responses, and TF‑IDF similarity to general capability corpora. BALIGN aggregates these features into a composite risk score to filter out high-risk preference samples, thereby preserving foundational capabilities while maintaining alignment gains with minimal computational overhead.
By Minsu Kim, Jianxun Lian, Xing Xie, Steven Euijong Whang
arXiv:2606. 30339v1 Announce Type: cross Abstract: Aligning large language models (LLMs) with diverse user preferences is a critical yet challenging task.
By Fuxiang Zhang, Pengcheng Wang, Chenran Li, Yi-Chen Li, Yuxin Chen, Lang Feng, Chenfeng Xu, Masayoshi Tomizuka, Bo An
arXiv:2509. 23982v2 Announce Type: replace-cross Abstract: Preference alignment is a critical step in making Large Language Models (LLMs) useful and aligned with (human) preferences.
By Lucio La Cava, Andrea Tagarelli
arXiv:2608.03446v2 Announce Type: replace
Abstract: Multilingual large language models (LLMs) have been shown to perform better on non-English classification tasks when the representations of the giv...
By Adnan Al Ali, Kathy H\"ammerl, Jind\v{r}ich Libovick\'y, Alexander Fraser
The paper introduces Comparison-based Preference Optimization (ComPO), a zeroth-order method that aligns large language models with human preferences using comparison oracles instead of direct differentiable loss optimization. It provides theoretical convergence guarantees for both offline and online variants under smoothness, gradient sparsity, and oracle compatibility assumptions, and establishes performance bounds under local coverage and in-distribution reward accuracy. Experiments on several LLMs (Mistral, Llama, Gemma-2, Qwen3, Gemma-3) show that ComPO outperforms existing direct alignment methods, achieving higher length-controlled win rates and diagnostics that suggest mitigation of likelihood displacement.
By Peter Chen, Xi Chen, Wotao Yin, Tianyi Lin
InfiFPO is a new preference optimization technique for implicit model fusion that replaces the reference model in Direct Preference Optimization with a fused source model, preserving probability information and avoiding complex vocabulary alignment. By incorporating probability clipping and max‑margin fusion, it aligns a pivot model with human preferences while distilling knowledge from multiple source models. Experiments on 11 benchmarks show that InfiFPO consistently outperforms existing fusion and preference optimization methods, raising Phi‑4’s average score from 79.95 to 83.33 and improving performance in mathematics, coding, and reasoning tasks.
By Yanggan Gu, Yuanyi Wang, Zhaoyi Yan, Yiming Zhang, Qi Zhou, Fei Wu, Hongxia Yang
Preference optimisation has proven effective for improving large language models but typically relies on costly human preference annotations. Extending these methods to morphologically rich, low-resource languages remains challenging because such annotations are scarce.
The paper introduces GAP-DPO, a method for personalizing large language models by selecting preference pairs based on gradient alignment with user utility. It formalizes personalized preference learning as a geometry‑aligned optimization problem, showing that off‑policy sampling can shift DPO updates from error correction to reinforcement when preference margins align with utility gradients. Experiments demonstrate that GAP‑DPO improves stylistic fidelity, preference alignment, and overall generation quality over standard DPO variants.
By Ruoming Jin, Xinyu Li, Hao Zhou, Jianfeng Zhu, Ruixin Guo, Feodor Dragan, Lei Xu, Haixun Wang, Yang Zhou