Fan Lai

Fan Lai

Auto-generated profile, based on publicly available information

Fan Lai

University of Illinois Urbana-Champaign, Champaign, USA

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
8 May 2023
Verified
Conference Paper
Open access
YWYiding Wang
Yiding Wang
DSDecang Sun
Decang Sun
KCKai Chen
Kai Chen
FLFan Lai
Fan Lai
MCMosharaf Chowdhury
Mosharaf Chowdhury

30 October 2023
Verified
Conference Paper
Open access
JLJiachen Liu
Jiachen Liu
FLFan Lai
Fan Lai
YDYinwei Dai
Yinwei Dai
AAAditya Akella
Aditya Akella
HMHarsha V. Madhyastha
Harsha V. Madhyastha
MCMosharaf Chowdhury
Mosharaf Chowdhury

12 October 2025
Verified
Conference Paper
Open access
YYYifan Yu
Yifan Yu
YGYu Gan
Yu Gan
NSNikhil Sarda
Nikhil Sarda
LTLillian Tsai
Lillian Tsai
JSJiaming Shen
Jiaming Shen
DCDavid Culler
David Culler

17 November 2025 in IEEE Transactions on Computers
Verified
Journal Article
Closed access
QLQinyi Luo
Qinyi Luo
PWPenghan Wang
Penghan Wang
WZWei Zhang
Wei Zhang
FLFan Lai
Fan Lai
JMJiachen Mao
Jiachen Mao
XQXuehai Qian
Xuehai Qian
Huge embedding tables in modern deep learning recommender models (DLRM) require prohibitively large memory during training and inference. This paper proposes FIITED, a system to automatically reduce the memory footprint via FIne-grained In-Training Embedding Dimension pruning. By leveraging the key insight that embedding vectors are not equally important, FIITED adaptively adjusts the dimension of each individual embedding vector during model training, assigning larger dimensions to more important embeddings while adapting to dynamic changes in data. We prioritize embedding dimensions with higher frequencies and gradients as more important. To enable efficient pruning of embeddings and their dimensions during model training, we propose an embedding storage system based on virtually-hashed physically-indexed hash tables. Experiments on two industry models and months of realistic datasets show that FIITED can reduce DLRM embedding size by more than 65% while preserving model quality, outperforming state-of-the-art in-training embedding pruning methods and increasing the reduction ratio by 1.3$\boldsymbol{\times}$ to 1.67$\boldsymbol{\times}$. On public datasets, FIITED can reduce the size of embedding tables by 2.2$\boldsymbol{\times}$ to 800$\boldsymbol{\times}$ with negligible accuracy drop, achieving 1.05$\boldsymbol{\times}$ to 12.5$\boldsymbol{\times}$ improvement in reduction ratio compared to baselines while improving model throughput.


11 October 2025
Verified
Conference Paper
Open access
YHYibo Huang
Yibo Huang
YQYiming Qiu
Yiming Qiu
ZYZhenning Yang
Zhenning Yang
YDYi Dai
Yi Dai
DWDingming Wu
Dingming Wu
ACAng Chen
Ang Chen