Submitted by lliutianc 7 OpenRubrics: Towards Scalable Synthetic Rubric Generation for Reward Modeling and LLM Alignment OpenRubrics 2