SIGNPOST-Bench: Benchmarking Text–Vision Conflict Resolution in Multimodal Large Language Models

SIGNPOST-Bench: Benchmarking Text–Vision Conflict Resolution in Multimodal Large Language Models | ResearchPod