Shaoyu Wu — accepted AI-conference papers

2026

D-CORE: Incentivizing Task Decomposition in Large Reasoning Models for Complex Tool Use

Bowen Xu, Shaoyu Wu, Hao Jiang, Kai Liu, Xin Chen, lulu hu +1

ICML 2026poster

Effective tool use and reasoning are essential capabilities for large reasoning models (LRMs) to address complex real-world problems. Through empirical analysis, we identify a prevalent "Lazy Reasoning" phenomenon, where LRMs frequently engage in repetitive and meaningless reflective reasoning. This…

Cited by 0 Source ↗Scholar ↗

2025

La RoSA: Enhancing LLM Efficiency via Layerwise Rotated Sparse Activation

Kai Liu, Bowen Xu, Shaoyu Wu, Xin Chen, Hao Zhou, Yongliang Tao +1

ICML 2025poster

Activation sparsity can reduce the computational overhead and memory transfers during the forward pass of Large Language Model (LLM) inference. Existing methods face limitations, either demanding time-consuming recovery training that hinders real-world adoption, or relying on empirical magnitude-bas…

Cited by 0 Source ↗PDF ↗Scholar ↗