From Outcomes to Processes: Guiding PRM Learning from ORM for Inference-Time Alignment

Open in new window