Fengxiang Bie — accepted AI-conference papers

2026

CARE: Covariance-Aware and Rank-Enhanced Decomposition for Enabling Multi-Head Latent Attention

Zhongzhu Zhou, Fengxiang Bie, Ziyan Chen, Zhenyu Zhang, Yibo Yang, Junxiong Wang +3

ICLR 2026poster

Converting pretrained attention modules such as *grouped-query attention* (GQA) into *multi-head latent attention* (MLA) can improve expressivity without increasing KV-cache cost, making it attractive for efficient inference. However, existing conversion methods typically apply naïve singular value…

Cited by 0 Source ↗Scholar ↗

2026

When RL Meets Adaptive Speculative Training: A Unified Training-Serving System

Fengxiang Bie, Junxiong Wang, Jisen Li, Zhongzhu Zhou, Chenfeng Xu, Zelei Shao +8

ICML 2026poster

Speculative decoding can significantly accelerate LLM serving, but its real-world benefits often erode due to training–serving mismatch and non-stationary traffic. Unlike previous systems that decouple speculator training from inference, we present a unified training–serving system, Aurora, that clo…

Cited by 0 Source ↗Scholar ↗