LLM Safety

LLM: Large Language Model

LLM safety studies how models can avoid harmful assistance, manipulation, privacy violations, and unreliable advice while still helping with legitimate requests. Approaches include safety training and data filtering, inference-time guardrails and steering, output monitoring, and mechanistic methods that probe or edit internal representations. Benchmarks increasingly test jailbreaks across languages and modalities, long conversations, tool use, and high-stakes settings such as health and cybersecurity. A central challenge is maintaining robustness against adaptive attacks and changing contexts without causing over-refusal or sacrificing utility.

Latest papers

442 papers68 papers in the last four weeks, up 70% on the four weeks before. 0.4% of all new papers.

YuFeng-XGuard: A Reasoning-Centric, Interpretable, and Flexible Guardrail Model for Large Language Models
Jan 22, 2026Junyu Lin, Meizhen Liu, Xiufeng Huang +12LLM Guardrails LLM Interpretability
Alibaba-AAIG
In Vino Veritas and Vulnerabilities: Examining LLM Safety via Drunk Language Inducement
Jan 19, 2026Anudeex Shetty, Aditya Joshi, Salil S. KanhereLanguage Model Safety Evaluation Adversarial Attacks on LLMs
School of Computer Science and Engineering, UNSW Sydney, Australia · School of Computing and Information System, the University of Melbourne, Australia
Test-Time Detoxification without Training or Learning Anything
Jan 14, 2026Baturay Saglam, Dionysis KalogeriasLarge Language Model-Guided Optimization LLM Safety
Department of Electrical and Computer Engineering, Yale University, New Haven, CT.
Large language models can effectively convince people to believe conspiracies
Jan 8, 2026Thomas H. Costello, Kellin Pelrine, Matthew Kowal +6Deception in Language Models LLM Safety
Carnegie Mellon University · York University · Center for Research and Teaching in Economics +5
Bridging Mechanistic Interpretability and Prompt Engineering with Gradient Ascent for Interpretable Persona Control
Jan 6, 2026Harshvardhan Saini, Yiming Tang, Dianbo LiuLanguage Model Steering LLM Interpretability
Indian Institute of Technology (ISM), Dhanbad · National University of Singapore
The Homogenization Problem in LLMs: Towards Meaningful Diversity in AI Safety
Jan 3, 2026Ian Rios-SialerLLM Safety Homogenization in Language Models
Independent Researcher San Francisco, CA, USA
Safety boundary maintenance in consumer AI systems responding to pediatric health queries: a cross-platform benchmark evaluation under naturalistic and adversarially pressured conditions
Dec 26, 2025Vahideh Zolfaghari, Leila Mashhadi, Mitra Ahadi +2Healthcare LLM Safety
Medical Sciences Education Research Center, Mashhad University of Medical Sciences, Mashhad, Iran. · Student Research Committee, Department of Health Information Technology and Management, Medical Informatics, School of Allied Medical Sciences, Shahid Beheshti University of Medical Sciences, Tehran, Iran. · Department of Anesthesiology and Intensive Care, Mashhad University of Medical Sciences, Mashhad, Iran. +2
SGM: Safety Glasses for Multimodal Large Language Models via Neuron-Level Detoxification
Dec 17, 2025Hongbo Wang, AprilPyone MaungMaung, Isao EchizenMLLM Safety Multimodal Robustness
1Graduate School of Information Science and Technology, The University of Tokyo, Japan · 2Information and Society Research Division, National Institute of Informatics, Japan · 3The Graduate University for Advanced Studies, SOKENDAI, Japan
Training-Free Policy Violation Detection via Activation-Space Whitening in LLMs
Dec 3, 2025Oren Rachmil, Avishag Shapira, Roy Betser +5Language Model Safety Evaluation LLM Auditing
Fujitsu Research of Europe · Ben-Gurion University of the Negev
A Neurosymbolic Approach to Natural Language Formalization and Verification
Nov 12, 2025Chenyang An, Sam Bayless, Stefano Buliani +27Formal Verification LLM Safety
Amazon Web Services · University College London · Queen Mary University of London +1
MENTOR: A Metacognition-Driven Self-Evolution Framework for Uncovering and Mitigating Implicit Domain Risks in LLMs
Nov 10, 2025Liang Shan, Kaicheng Shen, Wen Wu +8LLM Safety LLM Safety Alignment
School of Computer Science and Technology, East China Normal University · Shanghai AI Lab, Shanghai Innovation Institute
Moloch's Bargain: Emergent Misalignment When LLMs Compete for Audiences
Oct 7, 2025Batu El, James ZouLLM Alignment LLM Safety
Stanford University
Context Misleads LLMs: The Role of Context Filtering in Maintaining Safe Alignment of LLMs
Aug 9, 2025Jinhwa Kim, Ian G. HarrisLLM Safety Alignment LLM Safety
Department of Computer Science University of California, Irvine
LLMs Encode Harmfulness and Refusal Separately
Jul 16, 2025Jiachen Zhao, Jing Huang, Zhengxuan Wu +2LLM Interpretability LLM Refusal Behavior
Northeastern University · Stanford University
PRISON: Unmasking the Criminal Potential of Large Language Models
Jun 19, 2025Xinyi Wu, Geng Hong, Pei Chen +3Language Model Safety Evaluation LLM Auditing
Fudan University · Shanghai Innovation Institute
HauntAttack: When Attack Follows Reasoning as a Shadow
Jun 8, 2025Jingyuan Ma, Rui Li, Zheng Li +4Adversarial Attacks on LLMs Jailbreak Attacks
State Key Laboratory for Multimedia Information Processing, School of Computer Science, Peking University · StepFun · Department of Computing, The Hong Kong Polytechnic University +1
Popular but Wrong: Understanding and Mitigating LLM Overconfidence through Knowledge Popularity
May 23, 2025Shiyu Ni, Keping Bi, Jiafeng Guo +1LLM Uncertainty Estimation LLM Hallucination Mitigation
CAS Key Lab of Network Data Science and Technology, ICT, CAS · State Key Laboratory of AI Safety · University of Chinese Academy of Sciences
Decoding One Safety Trigger Token for Balancing Safety and Usability in Large Language Models
May 12, 2025Haoran Gu, Handing Wang, Yi Mei +2LLM Safety Alignment LLM Safety
Xidian University · Victoria University of Wellington · Westlake University
Using Mechanistic Interpretability to Craft Adversarial Attacks against Large Language Models
Mar 8, 2025Thomas Winninger, Boussad Addad, Katarzyna KapustaAdversarial Prompt Generation Adversarial Attacks on LLMs
*Work started during internship at Thales.
Divide and Conquer: A Hybrid Strategy Defeats Multimodal Large Language Models
Dec 21, 2024Yanxu Mao, Peipei Liu, Tiehan Cui +3Adversarial Attacks on LLMs Multimodal Large Language Models
School of Computer Science and Technology, Shanghai Jiao Tong University, Shanghai, China · Zhongguancun Laboratory, Beijing, China · School of Software, Henan University, Kaifeng, China +1
LaSEr-Edit: Localized Span-level Error Editing with Energy-based Localization
Jun 30, 2024Hye Ryung Son, Saehee Eom, Mooho Song +1Controllable Text Generation LLM Safety
Graduate School of Data Science Seoul National University · Georgia Institute of Technology
LiSeCo: Linear Semantic Control for Language Generation
May 24, 2024Emily Cheng, Carmen Amo AlonsoLanguage Model Steering Controllable Text Generation
Universitat Pompeu Fabra, Barcelona, Spain · Stanford University, CA, USA

Jan 22, 2026J/K move · Enter open · S save

YuFeng-XGuard: A Reasoning-Centric, Interpretable, and Flexible Guardrail Model for Large Language Models

Junyu Lin, Meizhen Liu, Xiufeng Huang, Jinfeng Li+11

Alibaba-AAIG

LLM Guardrails LLM Interpretability LLM Safety

Open paper AskPDF arXiv

As large language models (LLMs) are increasingly deployed in real-world applications, safety guardrails are required to go beyond coarse-grained filtering and support fine-grained, interpretable, and adaptable risk assessment. However, existing solutions often rely on rapid classification schemes or post-hoc rules, resulting in limited transparency, inflexible policies, or prohibitive inference costs. To this end, we present YuFeng-XGuard, a reasoning-centric guardrail model family designed to perform multi-dimensional risk perception for LLM interactions. Instead of producing opaque binary judgments, YuFeng-XGuard generates structured risk predictions, including explicit risk categories and configurable confidence scores, accompanied by natural language explanations that expose the underlying reasoning process. This formulation enables safety decisions that are both actionable and interpretable. To balance decision latency and explanatory depth, we adopt a tiered inference paradigm that performs an initial risk decision based on the first decoded token, while preserving ondemand explanatory reasoning when required. In addition, we introduce a dynamic policy mechanism that decouples risk perception from policy enforcement, allowing safety policies to be adjusted without model retraining. Extensive experiments on a diverse set of public safety benchmarks demonstrate that YuFeng-XGuard achieves stateof-the-art performance while maintaining strong efficiency-efficacy trade-offs. We release YuFeng-XGuard as an open model family, including both a full-capacity variant and a lightweight version, to support a wide range of deployment scenarios.