Collections

Fairness benchmark for multimodal LLMs: dataset, image perturbations, and results (paper: Fairness Failure Modes of Multimodal LLMs).