TRUST-VLM: Thorough Red-Teaming for Uncovering Safety Threats in Vision-Language Models

Devs

TRUST-VLM: Thorough Red-Teaming for Uncovering Safety Threats in Vision-Language Models | Read Paper on Bytez