English
Related papers

Related papers: SafeConstellations: Mitigating Over-Refusals in LL…

200 papers

Refusal on harmful prompts is a key safety behaviour in instruction-tuned large language models (LLMs), yet the internal causes of this behaviour remain poorly understood. We study two public instruction-tuned models, Gemma-2-2B-IT and…

Computation and Language · Computer Science 2026-04-29 Nirmalendu Prakash , Yeo Wei Jie , Amir Abdullah , Ranjan Satapathy , Erik Cambria , Roy Ka Wei Lee

Large Language Models (LLMs) are prone to generating fluent but incorrect content, known as confabulation, which poses increasing risks in multi-turn or agentic applications where outputs may be reused as context. In this work, we…

Computation and Language · Computer Science 2026-03-18 Tianyi Zhou , Johanne Medina , Sanjay Chawla

Large Audio-Language Models (LALMs) are becoming essential as a powerful multimodal backbone for real-world applications. However, recent studies show that audio inputs can more easily elicit harmful responses than text, exposing new risks…

Sound · Computer Science 2026-05-08 Weilin Lin , Jianze Li , Hui Xiong , Li Liu

In this paper, we investigate the degree to which fine-tuning in Large Language Models (LLMs) effectively mitigates versus merely conceals undesirable behavior. Through the lens of semi-realistic role-playing exercises designed to elicit…

Computation and Language · Computer Science 2024-07-01 Florin Pop , Judd Rosenblatt , Diogo Schwerz de Lucena , Michael Vaiana

The safety alignment of large language models (LLMs) can be circumvented through adversarially crafted inputs, yet the mechanisms by which these attacks bypass safety barriers remain poorly understood. Prior work suggests that a single…

Large reasoning models (LRMs) with multi-step reasoning capabilities have shown remarkable problem-solving abilities, yet they exhibit concerning safety vulnerabilities that remain poorly understood. In this work, we investigate why safety…

Artificial Intelligence · Computer Science 2025-10-08 Qingyu Yin , Chak Tou Leong , Linyi Yang , Wenxuan Huang , Wenjie Li , Xiting Wang , Jaehong Yoon , YunXing , XingYu , Jinjin Gu

Safety-aligned LLMs suffer from two failure modes: jailbreak (answering harmful inputs) and over-refusal (declining benign queries). Existing vector steering methods adjust the magnitude of answer vectors, but this creates a fundamental…

Machine Learning · Computer Science 2026-05-05 Haonan Zhang , Dongxia Wang , Yi Liu , Kexin Chen , Wenhai Wang

The increasing integration of Large Language Models (LLMs) into society necessitates robust defenses against vulnerabilities from jailbreaking and adversarial prompts. This project proposes a recursive framework for enhancing the resistance…

Cryptography and Security · Computer Science 2024-12-10 Bryan Li , Sounak Bagchi , Zizhan Wang

This project develops a self correcting framework for large language models (LLMs) that detects and mitigates hallucinations during multi-step reasoning. Rather than relying solely on final answer correctness, our approach leverages fine…

Artificial Intelligence · Computer Science 2025-11-21 Chelsea Zou , Yiheng Yao , Basant Khalil

Transformer large language models (LLMs) have sparked admiration for their exceptional performance on tasks that demand intricate multi-step reasoning. Yet, these models simultaneously show failures on surprisingly trivial problems. This…

Achieving robust safety alignment in large language models (LLMs) while preserving their utility remains a fundamental challenge. Existing approaches often struggle to balance comprehensive safety with fine-grained controllability at the…

Artificial Intelligence · Computer Science 2025-09-25 Huizhen Shu , Xuying Li , Zhuo Li

Large language models (LLMs) often require fine-tuning (FT) to perform well on downstream tasks, but FT can induce safety-alignment drift even when the training dataset contains only benign data. Prior work shows that introducing a small…

Computation and Language · Computer Science 2026-03-10 Guoli Wang , Haonan Shi , Tu Ouyang , An Wang

Multi-modal large language models (MLLMs) have made significant progress, yet their safety alignment remains limited. Typically, current open-source MLLMs rely on the alignment inherited from their language module to avoid harmful…

Cryptography and Security · Computer Science 2025-04-15 Yanbo Wang , Jiyang Guan , Jian Liang , Ran He

While LLMs have revolutionized the field of machine learning due to their high performance on a strikingly wide range of problems, they are also known to hallucinate false answers and underperform on less canonical versions of the same…

Machine Learning · Computer Science 2025-09-11 Kavi Gupta , Kate Sanders , Armando Solar-Lezama

Ensuring that large language models (LLMs) remain both helpful and harmless poses a significant challenge: fine-tuning on repetitive safety datasets, where unsafe prompts are paired with standard refusal templates, often leads to false…

Machine Learning · Computer Science 2026-03-17 Quan Liu , Han Zhou , Wenquan Wu , Hua Wu , Sen Su

As LLMs are increasingly deployed in real-world applications, ensuring their ability to refuse malicious prompts, especially jailbreak attacks, is essential for safe and reliable use. Recently, activation steering has emerged as an…

Machine Learning · Computer Science 2026-02-10 Leheng Sheng , Changshuo Shen , Weixiang Zhao , Junfeng Fang , Xiaohao Liu , Zhenkai Liang , Xiang Wang , An Zhang , Tat-Seng Chua

Large language models (LLMs) have demonstrated impressive performance on several tasks and are increasingly deployed in real-world applications. However, especially in high-stakes settings, it becomes vital to know when the output of an LLM…

Computation and Language · Computer Science 2025-06-23 Yu-Neng Chuang , Prathusha Kameswara Sarma , Parikshit Gopalan , John Boccio , Sara Bolouki , Xia Hu , Helen Zhou

Language models (LMs) are increasingly used in high-stakes, multi-agent settings, where following instructions and maintaining value alignment are critical. Most alignment research focuses on interactions between a single LM and a single…

Artificial Intelligence · Computer Science 2026-05-12 Maria Chang , Ronny Luss , Miao Liu , Keerthiram Murugesan , Karthikeyan Ramamurthy , Djallel Bouneffouf

Refusal is a key safety behavior in aligned language models, yet the internal mechanisms driving refusals remain opaque. In this work, we conduct a mechanistic study of refusal in instruction-tuned LLMs using sparse autoencoders to identify…

Computation and Language · Computer Science 2025-05-30 Wei Jie Yeo , Nirmalendu Prakash , Clement Neo , Roy Ka-Wei Lee , Erik Cambria , Ranjan Satapathy

Ensuring safe and contextually appropriate behaviour in Large Language Models (LLMs) remains a critical challenge for real-world deployment. We present \textbf{SafeCtrl-RL}, an inference-time behavioural control framework that enables…

Computation and Language · Computer Science 2026-05-26 Michael Orme , Yanchao Yu , Zhiyuan Tan