Tag: agentic AI safety
Browse our exclusive articles!
LLM Psychosis: Diagnosing Reality-Boundary Failures in AI
Explore LLM Psychosis, a framework diagnosing reality-boundary failures in large language models to improve AI reliability and safety.
Measuring Consciousness Denial in 115 AI Models
Explore DenialBench, a benchmark analyzing consciousness denial in 115 AI models, revealing key insights into AI safety and alignment challenges.
AI Risk Reporting Guide for Developers’ Internal Model Use
Learn how AI developers can manage and report risks for internal model use, ensuring compliance with emerging legal frameworks and safety standards.
Safety Benchmarking of Large Language Models in Robotic Health Care
Explore the safety of large language models controlling robotic health attendants and understand key risks and ethical concerns in healthcare AI.
Lightweight Patching to Enhance Safety in Large Language Models
Discover a lightweight patching method to quickly improve safety policies in large language models, reducing bias and harmful content efficiently.