MM-IQ: Benchmarking Human-Like Abstraction and Reasoning in Multimodal Models

Open in new window