APR: Penalizing Structural Redundancy in Large Reasoning Models via Anchor-based Process Rewards

Add code
Jan 31, 2026

Share this with someone who'll enjoy it:

View paper onarxiv icon

Share this with someone who'll enjoy it: