English
Related papers

Related papers: LLM Safety Alignment is Divergence Estimation in D…

200 papers

Aligning large language models (LLMs) with human preferences has become essential for safe and beneficial AI deployment. While Reinforcement Learning from Human Feedback (RLHF) established the dominant paradigm, a proliferation of…

Artificial Intelligence · Computer Science 2026-01-13 Tarun Raheja , Nilay Pochhi

Reinforcement Learning from Human Feedback (RLHF) aims to align language models (LMs) with human values by training reward models (RMs) on binary preferences and using these RMs to fine-tune the base LMs. Despite its importance, the…

Machine Learning · Computer Science 2024-08-21 Manon Revel , Matteo Cargnelutti , Tyna Eloundou , Greg Leppert

Large Language Models have shown impressive generative capabilities across diverse tasks, but their safety remains a critical concern. Existing post-training alignment methods, such as SFT and RLHF, reduce harmful outputs yet leave LLMs…

Cryptography and Security · Computer Science 2025-10-21 Zhengyue Zhao , Yingzi Ma , Somesh Jha , Marco Pavone , Patrick McDaniel , Chaowei Xiao

Fine-tuning large language models (LLMs) improves performance but introduces critical safety vulnerabilities: even minimal harmful data can severely compromise safety measures. We observe that perturbations orthogonal to the alignment…

Machine Learning · Computer Science 2026-01-08 Shuo Yang , Qihui Zhang , Yuyang Liu , Xiaojun Jia , Kunpeng Ning , Jiayu Yao , Jigang Wang , Hailiang Dai , Yibing Song , Li Yuan

The safety alignment ability of Vision-Language Models (VLMs) is prone to be degraded by the integration of the vision module compared to its LLM backbone. We investigate this phenomenon, dubbed as ''safety alignment degradation'' in this…

Computation and Language · Computer Science 2024-10-14 Qin Liu , Chao Shang , Ling Liu , Nikolaos Pappas , Jie Ma , Neha Anna John , Srikanth Doss , Lluis Marquez , Miguel Ballesteros , Yassine Benajiba

Safety alignment incurs safety tax that perturbs a large reasoning model's (LRM) general reasoning ability. Existing datasets used for safety alignment for an LRM are usually constructed by distilling safety reasoning traces and answers…

Artificial Intelligence · Computer Science 2026-02-03 Yingsha Xie , Tiansheng Huang , Enneng Yang , Rui Min , Wenjie Lu , Xiaochun Cao , Naiqiang Tan , Li Shen

Ensuring the safety of Large Language Models (LLMs) in diverse linguistic settings remains challenging, particularly for low-resource languages. Existing safety alignment methods are English-centric, limiting their effectiveness. We…

Computation and Language · Computer Science 2025-04-09 Isaac Lim , Shaun Khoo , Roy Ka-Wei Lee , Watson Chua , Jia Yi Goh , Jessica Foo

Merging Large Language Models (LLMs) is a cost-effective technique for combining multiple expert LLMs into a single versatile model, retaining the expertise of the original ones. However, current approaches often overlook the importance of…

Computation and Language · Computer Science 2024-06-21 Hasan Abed Al Kader Hammoud , Umberto Michieli , Fabio Pizzati , Philip Torr , Adel Bibi , Bernard Ghanem , Mete Ozay

As large language models (LLMs) become integral to various applications, ensuring both their safety and utility is paramount. Jailbreak attacks, which manipulate LLMs into generating harmful content, pose significant challenges to this…

Cryptography and Security · Computer Science 2025-02-10 Guobin Shen , Dongcheng Zhao , Yiting Dong , Xiang He , Yi Zeng

The demand for regulating potentially risky behaviors of large language models (LLMs) has ignited research on alignment methods. Since LLM alignment heavily relies on reward models for optimization or evaluation, neglecting the quality of…

Computation and Language · Computer Science 2024-10-01 Yan Liu , Xiaoyuan Yi , Xiaokang Chen , Jing Yao , Jingwei Yi , Daoguang Zan , Zheng Liu , Xing Xie , Tsung-Yi Ho

Human cognition, driven by complex neurochemical processes, oscillates between imagination and reality and learns to self-correct whenever such subtle drifts lead to hallucinations or unsafe associations. In recent years, LLMs have…

Computation and Language · Computer Science 2026-01-09 Sharanya Dasgupta , Arkaprabha Basu , Sujoy Nath , Swagatam Das

We investigate the efficacy of Large Language Models (LLMs) in detecting implicit and explicit hate speech, examining how models with minimal safety alignment (uncensored) compare with more heavily aligned (censored) counterparts in a…

Computation and Language · Computer Science 2026-05-05 Sanjeeevan Selvaganapathy , Mehwish Nasim

We demonstrate that preference optimization methods can effectively enhance LLM safety. Applying various alignment techniques to the Falcon 11B model using safety datasets, we achieve a significant boost in global safety score (from…

Machine Learning · Computer Science 2024-09-13 Reda Alami , Ali Khalifa Almansoori , Ahmed Alzubaidi , Mohamed El Amine Seddik , Mugariya Farooq , Hakim Hacid

Preference-based alignment methods (e.g., RLHF, DPO) typically optimize a single scalar objective, implicitly averaging over heterogeneous human preferences. In practice, systematic annotator and user-group disagreement makes mean-reward…

Machine Learning · Computer Science 2026-05-19 Mingxi Zou , Jiaxiang Chen , Junfan Li , Langzhang Liang , Qifan Wang , Xu Yinghui , Zenglin Xu

Safety alignment of large language models (LLMs) has been gaining increasing attention. However, current safety-aligned LLMs suffer from the fragile and imbalanced safety mechanisms, which can still be induced to generate unsafe responses,…

Computation and Language · Computer Science 2024-12-18 Weixiang Zhao , Yulin Hu , Zhuojun Li , Yang Deng , Jiahe Guo , Xingyu Sui , Yanyan Zhao , Bing Qin , Tat-Seng Chua , Ting Liu

Reinforcement Learning from Human Feedback (RLHF) is increasingly used to align large language models (LLMs) with human preferences. However, the effectiveness of RLHF in addressing underlying biases remains unclear. This study investigates…

Computation and Language · Computer Science 2025-03-13 Logan Barnhart , Reza Akbarian Bafghi , Stephen Becker , Maziar Raissi

Large Language Models (LLMs) acquire extensive knowledge and remarkable abilities from extensive text corpora, making them powerful tools for various applications. To make LLMs more usable, aligning them with human preferences is essential.…

Computation and Language · Computer Science 2024-10-21 Mozhi Zhang , Pengyu Wang , Chenkun Tan , Mianqiu Huang , Dong Zhang , Yaqian Zhou , Xipeng Qiu

While Reinforcement Learning from Human Feedback (RLHF) is widely used to align Large Language Models (LLMs) with human preferences, it typically assumes homogeneous preferences across users, overlooking diverse human values and minority…

Computation and Language · Computer Science 2025-10-28 Yijiang River Dong , Tiancheng Hu , Yinhong Liu , Ahmet Üstün , Nigel Collier

The widespread application of large language models (LLMs) raises increasing demands on ensuring safety or imposing constraints, such as reducing harmful content and adhering to predefined rules. While there have been several works studying…

Machine Learning · Computer Science 2026-02-13 Yihan Du , Seo Taek Kong , R. Srikant

Reinforcement learning from human feedback (RLHF) has emerged as a powerful technique to make large language models (LLMs) more capable in complex settings. RLHF proceeds as collecting human preference data, training a reward model on said…

Machine Learning · Computer Science 2024-02-05 Nathan Lambert , Roberto Calandra