English
Related papers

Related papers: In-Context Representation Hijacking

200 papers

Modern LLMs employ safety mechanisms that extend beyond surface-level input filtering to latent semantic representations and generation-time reasoning, enabling them to recover obfuscated malicious intent during inference and refuse…

Computation and Language · Computer Science 2026-03-18 Xiaobing Sun , Perry Lam , Shaohua Li , Zizhou Wang , Rick Siow Mong Goh , Yong Liu , Liangli Zhen

Large Language Models (LLMs) have demonstrated exceptional performance across various tasks, but their security vulnerabilities can be exploited by attackers to generate harmful content, causing adverse impacts across various societal…

Cryptography and Security · Computer Science 2025-12-17 Fan Yang

Large Language Models (LLMs) have demonstrated remarkable capabilities in performing tasks across various domains without needing explicit retraining. This capability, known as In-Context Learning (ICL), while impressive, exposes LLMs to a…

Computation and Language · Computer Science 2024-10-16 Bibek Upadhayay , Vahid Behzadan , Amin Karbasi

While large language models (LLMs) have demonstrated increasing power, they have also given rise to a wide range of harmful behaviors. As representatives, jailbreak attacks can provoke harmful or unethical responses from LLMs, even after…

Computation and Language · Computer Science 2024-03-01 Nan Xu , Fei Wang , Ben Zhou , Bang Zheng Li , Chaowei Xiao , Muhao Chen

While multilingual large language models generally perform adequately, and sometimes even rival English performance on high-resource languages (HRLs), they often significantly underperform on low-resource languages (LRLs). Among several…

Computation and Language · Computer Science 2025-10-09 Yilei Tu , Andrew Xue , Freda Shi

In-context learning (ICL) enables large language models (LLMs) to acquire new behaviors from the input sequence alone without any parameter updates. Recent studies have shown that ICL can surpass the original meaning learned in pretraining…

Machine Learning · Computer Science 2025-07-31 Yongyi Yang , Hidenori Tanaka , Wei Hu

Large language models (LLMs) are susceptible to a type of attack known as jailbreaking, which misleads LLMs to output harmful contents. Although there are diverse jailbreak attack strategies, there is no unified understanding on why some…

Computation and Language · Computer Science 2024-12-04 Yuping Lin , Pengfei He , Han Xu , Yue Xing , Makoto Yamada , Hui Liu , Jiliang Tang

Large Language Model (LLM) jailbreak refers to a type of attack aimed to bypass the safeguard of an LLM to generate contents that are inconsistent with the safe usage guidelines. Based on the insights from the self-attention computation…

Cryptography and Security · Computer Science 2025-02-10 Zhilong Wang , Haizhou Wang , Nanqing Luo , Lan Zhang , Xiaoyan Sun , Yebo Cao , Peng Liu

Paraphrasing of offensive content is a better alternative to content removal and helps improve civility in a communication environment. Supervised paraphrasers; however, rely heavily on large quantities of labelled data to help preserve…

Computation and Language · Computer Science 2024-06-11 Anirudh Som , Karan Sikka , Helen Gent , Ajay Divakaran , Andreas Kathol , Dimitra Vergyri

Large Language Models (LLMs) struggle with long-horizon tasks due to the "context bottleneck" and the "lost-in-the-middle" phenomenon, where accumulated noise from verbose environments degrades reasoning over multi-turn interactions. To…

Artificial Intelligence · Computer Science 2026-04-14 Xiaozhe Li , Tianyi Lyu , Yizhao Yang , Liang Shan , Siyi Yang , Ligao Zhang , Zhuoyi Huang , Qingwen Liu , Yang Li

Retrieval-Augmented Generation (RAG) systems enhance response credibility and traceability by displaying reference contexts, but this transparency simultaneously introduces a novel black-box attack vector. Existing document poisoning…

Computation and Language · Computer Science 2026-01-27 Runqi Sui

Large Language Models (LLMs) are widely deployed in applications that accept user-submitted content, such as uploaded documents or pasted text, for tasks like summarization and question answering. In this paper, we identify a new class of…

Cryptography and Security · Computer Science 2025-08-28 Zhuotao Lian , Weiyu Wang , Qingkui Zeng , Toru Nakanishi , Teruaki Kitasuka , Chunhua Su

Large Language Models (LLMs) increasingly employ alignment techniques to prevent harmful outputs. Despite these safeguards, attackers can circumvent them by crafting adversarial prompts. Predominant token-level optimization methods…

Computation and Language · Computer Science 2026-05-12 Jiawei Lian , Jianhong Pan , Lefan Wang , Yi Wang , Tairan Huang , Shaohui Mei , Lap-Pui Chau

Large Language Models (LLMs) have demonstrated great capabilities in natural language understanding and generation, largely attributed to the intricate alignment process using human feedback. While alignment has become an essential training…

Computation and Language · Computer Science 2024-09-04 Bocheng Chen , Hanqing Guo , Guangjing Wang , Yuanda Wang , Qiben Yan

Writing effective prompts for large language models (LLM) can be unintuitive and burdensome. In response, services that optimize or suggest prompts have emerged. While such services can reduce user effort, they also introduce a risk: the…

Cryptography and Security · Computer Science 2025-03-03 Weiran Lin , Anna Gerchanovsky , Omer Akgul , Lujo Bauer , Matt Fredrikson , Zifan Wang

Large Language Models (LLMs) are increasingly vulnerable to a sophisticated form of adversarial prompting known as camouflaged jailbreaking. This method embeds malicious intent within seemingly benign language to evade existing safety…

Cryptography and Security · Computer Science 2025-09-09 Youjia Zheng , Mohammad Zandsalimy , Shanu Sushmita

Large language models (LLMs) have transformed the development of embodied intelligence. By providing a few contextual demonstrations, developers can utilize the extensive internal knowledge of LLMs to effortlessly translate complex tasks…

Artificial Intelligence · Computer Science 2024-08-07 Aishan Liu , Yuguang Zhou , Xianglong Liu , Tianyuan Zhang , Siyuan Liang , Jiakai Wang , Yanjun Pu , Tianlin Li , Junqi Zhang , Wenbo Zhou , Qing Guo , Dacheng Tao

Large language models (LLMs) are equipped with safety mechanisms to detect and block harmful queries, yet current alignment approaches primarily focus on overtly dangerous content and overlook more subtle threats. However, users can often…

Computation and Language · Computer Science 2026-01-01 Shenzhe Zhu

Large language models (LLMs) have succeeded significantly in various applications but remain susceptible to adversarial jailbreaks that void their safety guardrails. Previous attempts to exploit these vulnerabilities often rely on high-cost…

Machine Learning · Computer Science 2024-12-02 Xuan Li , Zhanke Zhou , Jianing Zhu , Jiangchao Yao , Tongliang Liu , Bo Han

Large Language Models (LLMs) are powerful tools with profound societal impacts, yet their ability to generate responses to diverse and uncontrolled inputs leaves them vulnerable to adversarial attacks. While existing defenses often struggle…

Computation and Language · Computer Science 2025-12-30 Samuel Simko , Mrinmaya Sachan , Bernhard Schölkopf , Zhijing Jin