Donglei Wu — accepted AI-conference papers

2026

RCMoE: A Communication-Efficient Random Compression Framework for Resource-Constrained Mixture-of-Experts Training

Donglei Wu, Xiao Cai, Jinglei Tan, Jinda Jia, Guangming Tan, Dingwen Tao +2

AAAI 2026technical

Mixture-of-Experts (MoE) architecture with experts parallelism scales LLMs efficiently by activating only a subset of experts per input, avoiding proportional training costs. However, the intensive and heterogeneous communication substantially hinders the efficiency and scalability of MoE training i

Cited by 0 Source ↗PDF ↗Scholar ↗

2022

SmartIdx: Reducing Communication Cost in Federated Learning by Exploiting the CNNs Structures

Donglei Wu, Xiangyu Zou, Shuyu Zhang, Haoyu Jin, Wen Xia, Binxing Fang

AAAI 2022technical

Top-k sparsification method is popular and powerful forreducing the communication cost in Federated Learning(FL). However, according to our experimental observation, it spends most of the total communication cost on the index of the selected parameters (i.e., their position informa-tion), which is i…

Cited by 23 Source ↗PDF ↗code ↗Scholar ↗