English
Related papers

Related papers: Metaphor Is Not All Attention Needs

200 papers

Large Language Models (LLMs) have evolved into Multimodal Large Language Models (MLLMs), significantly enhancing their capabilities by integrating visual information and other types, thus aligning more closely with the nature of human…

Cryptography and Security · Computer Science 2025-06-03 Youze Wang , Wenbo Hu , Yinpeng Dong , Jing Liu , Hanwang Zhang , Richang Hong

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) face a significant threat from multi-turn jailbreak attacks, where adversaries progressively steer conversations to elicit harmful outputs. However, the practical effectiveness of existing attacks is undermined…

Cryptography and Security · Computer Science 2026-01-12 Songze Li , Ruishi He , Xiaojun Jia , Jun Wang , Zhihui Fu

Robust alignment guardrails for large language models (LLMs) are becoming increasingly important with their widespread application. In contrast to previous studies, we demonstrate that inference-time activation interventions can bypass…

Computation and Language · Computer Science 2025-08-26 Paul Darm , Annalisa Riccardi

Large language models (LLMs) remain vulnerable to sophisticated prompt engineering attacks that exploit contextual framing to bypass safety mechanisms, posing significant risks in cybersecurity applications. We introduce Jailbreak Mimicry,…

Cryptography and Security · Computer Science 2025-10-28 Pavlos Ntais

Large-scale pretrained language models are the major driving force behind recent improvements in performance on the Winograd Schema Challenge, a widely employed test of common sense reasoning ability. We show, however, with a new diagnostic…

Computation and Language · Computer Science 2020-05-08 Mostafa Abdou , Vinit Ravishankar , Maria Barrett , Yonatan Belinkov , Desmond Elliott , Anders Søgaard

As deep learning advances, Large Language Models (LLMs) and their multimodal counterparts, Multimodal Large Language Models (MLLMs), have shown exceptional performance in many real-world tasks. However, MLLMs face significant security…

Cryptography and Security · Computer Science 2024-10-23 Fenghua Weng , Yue Xu , Chengyan Fu , Wenjie Wang

Despite efforts to align large language models to produce harmless responses, they are still vulnerable to jailbreak prompts that elicit unrestricted behaviour. In this work, we investigate persona modulation as a black-box jailbreaking…

Computation and Language · Computer Science 2023-11-27 Rusheb Shah , Quentin Feuillade--Montixi , Soroush Pour , Arush Tagade , Stephen Casper , Javier Rando

The vulnerability of Vision Large Language Models (VLLMs) to jailbreak attacks appears as no surprise. However, recent defense mechanisms against these attacks have reached near-saturation performance on benchmark evaluations, often with…

Cryptography and Security · Computer Science 2025-03-07 Yangyang Guo , Fangkai Jiao , Liqiang Nie , Mohan Kankanhalli

This paper proposes a jailbreaking prompt detection method for large language models (LLMs) to defend against jailbreak attacks. Although recent LLMs are equipped with built-in safeguards, it remains possible to craft jailbreaking prompts…

Cryptography and Security · Computer Science 2026-05-12 Zheng Lin , Zhenxing Niu , Haoxuan Ji , Yuzhe Huang , Haichang Gao

This paper provides a systematic survey of jailbreak attacks and defenses on Large Language Models (LLMs) and Vision-Language Models (VLMs), emphasizing that jailbreak vulnerabilities stem from structural factors such as incomplete training…

Cryptography and Security · Computer Science 2026-01-08 Zejian Chen , Chaozhuo Li , Chao Li , Xi Zhang , Litian Zhang , Yiming He

The capacity of large language models (LLMs) to generate honest, harmless, and helpful responses heavily relies on the quality of user prompts. However, these prompts often tend to be brief and vague, thereby significantly limiting the full…

Computation and Language · Computer Science 2025-07-01 Xiaohua Wang , Zisu Huang , Feiran Zhang , Zhibo Xu , Cenyuan Zhang , Qi Qian , Xiaoqing Zheng , Xuanjing Huang

We identify a structural weakness in current large language model (LLM) alignment: modern refusal mechanisms are fail-open. While existing approaches encode refusal behaviors across multiple latent features, suppressing a single dominant…

Machine Learning · Computer Science 2026-02-20 Zachary Coalson , Beth Sohler , Aiden Gabriel , Sanghyun Hong

The integration of additional modalities increases the susceptibility of large vision-language models (LVLMs) to safety risks, such as jailbreak attacks, compared to their language-only counterparts. While existing research primarily…

Computation and Language · Computer Science 2025-06-24 Yilei Jiang , Xinyan Gao , Tianshuo Peng , Yingshui Tan , Xiaoyong Zhu , Bo Zheng , Xiangyu Yue

Uncovering the mechanisms behind "jailbreaks" in large language models (LLMs) is crucial for enhancing their safety and reliability, yet these mechanisms remain poorly understood. Existing studies predominantly analyze jailbreak prompts by…

Machine Learning · Computer Science 2026-02-06 Licheng Pan , Yunsheng Lu , Jiexi Liu , Jialing Tao , Haozhe Feng , Hui Xue , Zhixuan Chu , Kui Ren

Large language models (LLMs) pose significant risks due to the potential for generating harmful content or users attempting to evade guardrails. Existing studies have developed LLM-based guard models designed to moderate the input and…

Cryptography and Security · Computer Science 2025-02-25 Hongfu Liu , Hengguan Huang , Xiangming Gu , Hao Wang , Ye Wang

The safety alignment of large language models (LLMs) remains vulnerable, as their initial behavior can be easily jailbroken by even relatively simple attacks. Since infilling a fixed template between the input instruction and initial model…

Computation and Language · Computer Science 2025-06-05 Chak Tou Leong , Qingyu Yin , Jian Wang , Wenjie Li

Most commonly used language models (LMs) are instruction-tuned and aligned using a combination of fine-tuning and reinforcement learning, causing them to refuse users requests deemed harmful by the model. However, jailbreak prompts can…

Computation and Language · Computer Science 2025-07-02 Aryan Shrivastava , Ari Holtzman

AI systems are rapidly advancing in capability, and frontier model developers broadly acknowledge the need for safeguards against serious misuse. However, this paper demonstrates that fine-tuning, whether via open weights or closed…

Cryptography and Security · Computer Science 2025-09-23 Brendan Murphy , Dillon Bowen , Shahrad Mohammadzadeh , Tom Tseng , Julius Broomfield , Adam Gleave , Kellin Pelrine

While Large Language Models (LLMs) have shown significant advancements in performance, various jailbreak attacks have posed growing safety and ethical risks. Malicious users often exploit adversarial context to deceive LLMs, prompting them…

Cryptography and Security · Computer Science 2025-08-15 Jinhwa Kim , Ian G. Harris