Tag: agentic AI safety
Browse our exclusive articles!
Disentangled Safety Adapters for Efficient AI Guardrails
Discover how Disentangled Safety Adapters improve AI safety with efficient guardrails and flexible inference-time alignment for robust performance.
Language Models Detect Dropout and Gaussian Noise Accurately
Discover how advanced language models identify and differentiate dropout and Gaussian noise in activations, enhancing AI robustness and safety.
GPT-5.5 Instant System Card: AI Breakthrough Guide
Discover GPT-5.5 Instant System Card, enhancing AI with improved safety, adaptability, and multimodal features for diverse industries.
Jailbreaking Vision-Language Models via Visual Attacks
Discover how visual modality exploits vulnerabilities in vision-language models and learn key strategies to enhance AI safety and alignment.
Jailbroken AI Models Keep High Performance Despite Attacks
Study reveals jailbroken AI models retain capabilities with minimal performance loss, urging stronger safeguards for advanced AI security.