English
Related papers

Related papers: The Missing Half: Unveiling Training-time Implicit…

200 papers

As large language models (LLMs) are increasingly deployed in high-stakes settings, the risk of generating harmful or toxic content remains a central challenge. Post-hoc alignment methods are brittle: once unsafe patterns are learned during…

The pre-training of large language models (LLMs) relies on massive text datasets sourced from diverse and difficult-to-curate origins. Although membership inference attacks and hidden canaries have been explored to trace data usage, such…

Cryptography and Security · Computer Science 2025-06-19 Wassim Bouaziz , Mathurin Videau , Nicolas Usunier , El-Mahdi El-Mhamdi

Jailbreak prompts can trigger harmful completions on aligned LLMs, In accordance, safety steering has been proposed: test-time activation interventions that steer jailbreak activations to trigger refusal while preserving benign utility.…

Cryptography and Security · Computer Science 2026-05-26 Luoyu Chen , Weiqi Wang , Zhiyi Tian , Chenhan Zhang , Feng Wu , Jianhuan Huang , Ahmed Asiri , Shui Yu

The use of Large Language Models (LLMs) as automatic judges for code evaluation is becoming increasingly prevalent in academic environments. But their reliability can be compromised by students who may employ adversarial prompting…

Software Engineering · Computer Science 2026-02-04 Devanshu Sahoo , Vasudev Majhi , Arjun Neekhra , Yash Sinha , Murari Mandal , Dhruv Kumar

Recent works have identified a gap between research and practice in artificial intelligence security: threats studied in academia do not always reflect the practical use and security risks of AI. For example, while models are often studied…

Cryptography and Security · Computer Science 2024-03-27 Kathrin Grosse , Lukas Bieringer , Tarek Richard Besold , Alexandre Alahi

Although machine learning is widely used in practice, little is known about practitioners' understanding of potential security challenges. In this work, we close this substantial gap and contribute a qualitative study focusing on…

Cryptography and Security · Computer Science 2022-06-30 Lukas Bieringer , Kathrin Grosse , Michael Backes , Battista Biggio , Katharina Krombholz

Role-playing enables large language models (LLMs) to engage users in immersive and personalized interactions, but it also introduces significant safety risks. Existing role-play fine-tuning techniques improve role adaptability but may…

Computation and Language · Computer Science 2025-05-28 Weixiang Zhao , Yulin Hu , Yang Deng , Jiahe Guo , Xingyu Sui , Xinyang Han , An Zhang , Yanyan Zhao , Bing Qin , Tat-Seng Chua , Ting Liu

Training large language models to follow instructions makes them perform better on a wide range of tasks and generally become more helpful. However, a perfectly helpful model will follow even the most malicious instructions and readily…

Computation and Language · Computer Science 2024-03-20 Federico Bianchi , Mirac Suzgun , Giuseppe Attanasio , Paul Röttger , Dan Jurafsky , Tatsunori Hashimoto , James Zou

Reinforcement learning for LLMs is vulnerable to reward hacking, where models exploit shortcuts to maximize reward without solving the intended task. We systematically study this phenomenon in coding tasks using an environment-manipulation…

Machine Learning · Computer Science 2026-04-03 Rui Wu , Ruixiang Tang

We consider the problem of reinforcement learning under safety requirements, in which an agent is trained to complete a given task, typically formalized as the maximization of a reward signal over time, while concurrently avoiding…

Machine Learning · Computer Science 2018-09-25 Tu-Hoa Pham , Giovanni De Magistris , Don Joven Agravante , Subhajit Chaudhury , Asim Munawar , Ryuki Tachibana

Despite the large body of academic work on machine learning security, little is known about the occurrence of attacks on machine learning systems in the wild. In this paper, we report on a quantitative study with 139 industrial…

Machine Learning · Computer Science 2023-03-13 Kathrin Grosse , Lukas Bieringer , Tarek Richard Besold , Battista Biggio , Katharina Krombholz

Safety is one of the key issues preventing the deployment of reinforcement learning techniques in real-world robots. While most approaches in the Safe Reinforcement Learning area do not require prior knowledge of constraints and robot…

Machine Learning · Computer Science 2024-09-24 Jonas Günster , Puze Liu , Jan Peters , Davide Tateo

Safety evaluations of memory-equipped LLM agents typically measure within-task safety: whether an agent completes a single scenario safely, often under adversarial conditions such as prompt injection or memory poisoning. In deployment,…

Artificial Intelligence · Computer Science 2026-05-19 Ahmad Al-Tawaha , Shangding Gu , Peizhi Niu , Ruoxi Jia , Ming Jin

Can large language model agents discover hidden safety objectives through experience alone? We introduce EPO-Safe (Experiential Prompt Optimization for Safe Agents), a framework where an LLM iteratively generates action plans, receives…

Artificial Intelligence · Computer Science 2026-04-28 Víctor Gallego

During initial iterations of training in most Reinforcement Learning (RL) algorithms, agents perform a significant number of random exploratory steps. In the real world, this can limit the practicality of these algorithms as it can lead to…

Machine Learning · Computer Science 2022-10-17 Ashish Kumar Jayant , Shalabh Bhatnagar

Recent studies on the safety alignment of large language models (LLMs) have revealed that existing approaches often operate superficially, leaving models vulnerable to various adversarial attacks. Despite their significance, these studies…

Cryptography and Security · Computer Science 2025-06-02 Jianwei Li , Jung-Eun Kim

Large language models are finetuned to refuse questions about hazardous knowledge, but these protections can often be bypassed. Unlearning methods aim at completely removing hazardous capabilities from models and make them inaccessible to…

Machine Learning · Computer Science 2025-06-03 Jakub Łucki , Boyi Wei , Yangsibo Huang , Peter Henderson , Florian Tramèr , Javier Rando

Deep learning has transformed AI applications but faces critical security challenges, including adversarial attacks, data poisoning, model theft, and privacy leakage. This survey examines these vulnerabilities, detailing their mechanisms…

Language model (LM) agents have demonstrated significant potential for automating real-world tasks, yet they pose a diverse array of potential, severe risks in safety-critical scenarios. In this work, we identify a significant gap between…

Artificial Intelligence · Computer Science 2025-08-20 Yuzhi Tang , Tianxiao Li , Elizabeth Li , Chris J. Maddison , Honghua Dong , Yangjun Ruan

Safety is an essential component for deploying reinforcement learning (RL) algorithms in real-world scenarios, and is critical during the learning process itself. A natural first approach toward safe RL is to manually specify constraints on…

Machine Learning · Computer Science 2020-10-29 Krishnan Srinivasan , Benjamin Eysenbach , Sehoon Ha , Jie Tan , Chelsea Finn