MM-IFEval
Emerging2papers using it
2025first seen
MM-IFEval is a challenging and diverse multi-modal instruction-following benchmark that evaluates the performance of Multi-modal Large Language Models (MLLMs) using both compose-level and perception-level constraints.