English
Related papers

Related papers: Before the Last Token: Diagnosing Final-Token Safe…

200 papers

With the growing deployment of LLMs in daily applications like chatbots and content generation, efforts to ensure outputs align with human values and avoid harmful content have intensified. However, increasingly sophisticated jailbreak…

Cryptography and Security · Computer Science 2025-06-10 Cheng Qian , Hainan Zhang , Lei Sha , Zhiming Zheng

Latent-space monitors aim to detect undesirable behaviours in Large Language Models by leveraging their internal representations rather than relying solely on black-box outputs. These methods have shown promise in identifying behaviours…

Machine Learning · Computer Science 2026-02-27 Rohan Gupta , Erik Jenner

Despite advances in improving large language model (LLM) to refuse to answer malicious instructions, widely used LLMs remain vulnerable to jailbreak attacks where attackers generate instructions with distributions differing from safety…

Computation and Language · Computer Science 2025-09-05 Sheng Liu , Qiang Sheng , Danding Wang , Yang Li , Guang Yang , Juan Cao

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

Backdoor attacks pose a critical threat by embedding hidden triggers into inputs, causing models to misclassify them into target labels. While extensive research has focused on mitigating these attacks in object recognition models through…

Computer Vision and Pattern Recognition · Computer Science 2025-04-09 Kyle Stein , Andrew Arash Mahyari , Guillermo Francia , Eman El-Sheikh

As large language models (LLMs) advance, ensuring AI safety and alignment is paramount. One popular approach is prompt guards, lightweight mechanisms designed to filter malicious queries while being easy to implement and update. In this…

Machine Learning · Computer Science 2025-10-08 Jaiden Fairoze , Sanjam Garg , Keewoo Lee , Mingyuan Wang

Guardrail models (a.k.a. safety checkers) are widely deployed to screen user inputs before they reach large language models (LLMs), serving as a primary defense against prompt injection attacks. Due to strict context constraints, these…

Cryptography and Security · Computer Science 2026-05-25 Yuanbo Zhou , Changjia Zhu , Junyu Wang , Xu He , Yan Zhai , Kun Sun , Mingkui Wei , Junjie Xiong

As Large Language Models (LLMs) of Prompt Jailbreaking are getting more and more attention, it is of great significance to raise a generalized research paradigm to evaluate attack strengths and a basic model to conduct subtler experiments.…

Cryptography and Security · Computer Science 2024-04-15 Tianyu Zhang , Zixuan Zhao , Jiaqi Huang , Jingyu Hua , Sheng Zhong

Large language models remain vulnerable to jailbreak attacks, yet we still lack a systematic understanding of how jailbreak success scales with attacker effort across methods, model families, and harm types. We initiate a scaling-law…

Machine Learning · Computer Science 2026-03-20 Xiangwen Wang , Ananth Balashankar , Varun Chandrasekaran

Recently, Large Language Models (LLMs) have garnered significant attention for their exceptional natural language processing capabilities. However, concerns about their trustworthiness remain unresolved, particularly in addressing…

Computation and Language · Computer Science 2025-01-29 Yue Huang , Jingyu Tang , Dongping Chen , Bingda Tang , Yao Wan , Lichao Sun , Philip S. Yu , Xiangliang Zhang

Language models (LMs) are trained over sequences of tokens, whereas users interact with LMs via text. This mismatch gives rise to the partial token problem, which occurs when a user ends their prompt in the middle of the expected…

Computation and Language · Computer Science 2026-02-04 Hao Xu , Alisa Liu , Jonathan Hayase , Yejin Choi , Noah A. Smith

Large Language Models (LLMs) are increasingly being integrated into services such as ChatGPT to provide responses to user queries. To mitigate potential harm and prevent misuse, there have been concerted efforts to align the LLMs with human…

Cryptography and Security · Computer Science 2024-12-30 Xiaomeng Hu , Pin-Yu Chen , Tsung-Yi Ho

Red-teaming has been a widely adopted way to evaluate the harmfulness of Large Language Models (LLMs). It aims to jailbreak a model's safety behavior to make it act as a helpful agent disregarding the harmfulness of the query. Existing…

Computation and Language · Computer Science 2023-11-14 Rishabh Bhardwaj , Soujanya Poria

Large language model (LLM) systems increasingly power everyday AI applications such as chatbots, computer-use assistants, and autonomous robots, where performance often depends on manually well-crafted prompts. LLM-based prompt optimizers…

Machine Learning · Computer Science 2026-01-14 Andrew Zhao , Reshmi Ghosh , Vitor Carvalho , Emily Lawton , Keegan Hines , Gao Huang , Jack W. Stokes

Monitoring large language models' (LLMs) activations is an effective way to detect harmful requests before they lead to unsafe outputs. However, traditional safety monitors often require the same amount of compute for every query. This…

Machine Learning · Computer Science 2026-04-22 James Oldfield , Philip Torr , Ioannis Patras , Adel Bibi , Fazl Barez

Large Language Models (LLMs) are widely deployed in real-world systems. Given their broader applicability, prompt engineering has become an efficient tool for resource-scarce organizations to adopt LLMs for their own purposes. At the same…

Cryptography and Security · Computer Science 2026-02-27 Piyush Jaiswal , Aaditya Pratap , Shreyansh Saraswati , Harsh Kasyap , Somanath Tripathy

Safety fine-tuning of language models typically requires a curated adversarial dataset. We take a different approach: score each candidate prompt's difficulty by how often the target model's own rollouts are judged harmful, then fine-tune…

Machine Learning · Computer Science 2026-05-06 Prakhar Gupta , Garv Shah , Donghua Zhang

While reasoning large language models (LLMs) demonstrate remarkable performance across various tasks, they also contain notable security vulnerabilities. Recent research has uncovered a "thinking-stopped" vulnerability in DeepSeek-R1, where…

Cryptography and Security · Computer Science 2025-04-30 Yu Cui , Yujun Cai , Yiwei Wang

We show that even the most recent safety-aligned LLMs are not robust to simple adaptive jailbreaking attacks. First, we demonstrate how to successfully leverage access to logprobs for jailbreaking: we initially design an adversarial prompt…

Cryptography and Security · Computer Science 2025-04-21 Maksym Andriushchenko , Francesco Croce , Nicolas Flammarion

In this study, we disclose a worrying new vulnerability in Large Language Models (LLMs), which we term \textbf{involuntary jailbreak}. Unlike existing jailbreak attacks, this weakness is distinct in that it does not involve a specific…

Cryptography and Security · Computer Science 2025-12-30 Yangyang Guo , Yangyan Li , Mohan Kankanhalli