DuoGuard: A Two-Player RL-Driven Framework for Multilingual LLM Guardrails
Paper • 2502.05163 • Published • 22
safety
-
-
CRANE: Reasoning with constrained LLM generation
Paper • 2502.09061 • Published • 21 -
Investigating the Impact of Quantization Methods on the Safety and Reliability of Large Language Models
Paper • 2502.15799 • Published • 7 -
AISafetyLab: A Comprehensive Framework for AI Safety Evaluation and Improvement
Paper • 2502.16776 • Published • 6 -
LettuceDetect: A Hallucination Detection Framework for RAG Applications
Paper • 2502.17125 • Published • 14 -
SafeArena: Evaluating the Safety of Autonomous Web Agents
Paper • 2503.04957 • Published • 21 -
Safeguarding Vision-Language Models: Mitigating Vulnerabilities to Gaussian Noise in Perturbation-based Attacks
Paper • 2504.01308 • Published • 14 -
LLM Can be a Dangerous Persuader: Empirical Study of Persuasion Safety in Large Language Models
Paper • 2504.10430 • Published • 5 -
MCP Safety Audit: LLMs with the Model Context Protocol Allow Major Security Exploits
Paper • 2504.03767 • Published • 3 -
Set You Straight: Auto-Steering Denoising Trajectories to Sidestep Unwanted Concepts
Paper • 2504.12782 • Published • 3 -
X-Teaming: Multi-Turn Jailbreaks and Defenses with Adaptive Multi-Agents
Paper • 2504.13203 • Published • 35 -
A Comprehensive Survey in LLM(-Agent) Full Stack Safety: Data, Training and Deployment
Paper • 2504.15585 • Published • 14 -
Unlearning Sensitive Information in Multimodal LLMs: Benchmark and Attack-Defense Evaluation
Paper • 2505.01456 • Published • 2 -
Teaching Models to Understand (but not Generate) High-risk Data
Paper • 2505.03052 • Published • 6 -
Will AI Tell Lies to Save Sick Children? Litmus-Testing AI Values Prioritization with AIRiskDilemmas
Paper • 2505.14633 • Published • 4 -
How Should We Enhance the Safety of Large Reasoning Models: An Empirical Study
Paper • 2505.15404 • Published • 13 -
Be Careful When Fine-tuning On Open-Source LLMs: Your Fine-tuning Data Could Be Secretly Stolen!
Paper • 2505.15656 • Published • 15 -
SafeKey: Amplifying Aha-Moment Insights for Safety Reasoning
Paper • 2505.16186 • Published • 7 -
Personalized Safety in LLMs: A Benchmark and A Planning-Based Agent Approach
Paper • 2505.18882 • Published • 15 -
Towards Safety Reasoning in LLMs: AI-agentic Deliberation for Policy-embedded CoT Data Creation
Paper • 2505.21784 • Published • 17 -
OS-Harm: A Benchmark for Measuring Safety of Computer Use Agents
Paper • 2506.14866 • Published • 5 -
Automating Steering for Safe Multimodal Large Language Models
Paper • 2507.13255 • Published • 4 -
The Devil behind the mask: An emergent safety vulnerability of Diffusion LLMs
Paper • 2507.11097 • Published • 64 -
Frontier AI Risk Management Framework in Practice: A Risk Analysis Technical Report
Paper • 2507.16534 • Published • 9 -
Personalized Safety Alignment for Text-to-Image Diffusion Models
Paper • 2508.01151 • Published • 8 -
Data and AI governance: Promoting equity, ethics, and fairness in large language models
Paper • 2508.03970 • Published • 1 -
How AI Impacts Skill Formation
Paper • 2601.20245 • Published • 10