Tag
The Open MM-RL dataset, trending #1 on Hugging Face, offers PhD-level STEM problems with deterministic grading for multimodal RL training, including complex visual tasks double-vetted by domain specialists.