DRP: Distilled Reasoning Pruning with Mathematical Skill-aware Step Decomposition for Efficient Large Reasoning Models.

Yuxuan Jiang, Dawei Li, Francis Ferraro

Browse the full ACL paper archive.