ACL 2026: Huawei Taylor Lab Proposes SHAPE, Adding a Reasoning Tax to LLM Inference
ACL 2026|答得更准还写得更短?华为泰勒实验室提出SHAPE,给LLM推理装了个「推理税」
Huawei Taylor Lab, Peking University, and Shanghai University of Finance and Economics proposed SHAPE, accepted by ACL 2026, with about 3% average accuracy gain. It uses entropy segmentation, short rollouts for potential estimation, dynamic length discounts, and token-level credit assignment, cutting token use by about 30%. The key mechanism is a reasoning tax: long high-potential late-stage segments are penalized to reduce verbose confirmation loops.
Why it matters: Huawei's Taylor Lab and others propose SHAPE, which segments by entropy and probes short spans to estimate potential, then discounts long late sections so models answer math problems more accurately and more briefly.