Generative RLHF-V: Learning Principles from Multi-modal Human Preference

Re-assign community