Submitted by lliutianc 8 OpenRubrics: Towards Scalable Synthetic Rubric Generation for Reward Modeling and LLM Alignment OpenRubrics 2