作者位次 #3/4ACL 2026Findings· Findings of ACL
When Internalization Fails: Finding Better Targets for Reasoning Compression
Mourad Heddaya · Manley Roberts · Rohan Wadhawan · Chenhao Tan
摘要
Reasoning language models generate long reasoning traces that increase latency and cost. We study how to shorten these traces while preserving accuracy on competition-level mathematics. In a teacher-student distillation setup, we compare three approaches: (i) inference-time truncation after the f...