About

Wei Feng is a Software Engineer at Meta working on PyTorch distributed training

FSDP2: parameters, gradients, and optimizer states sharded across ranks, and support for float8 all-gather, and LoRA/QLoRA
FlexShard: pluggable sharding placements for Muon optimizer and block-wise quantization
TorchTitan: PyTorch-native, one-stop platform for production-ready LLM pretraining

Before that Wei worked on recommender systems at Meta and Hulu

Reels: user-to-item and item-to-item retrieval, improve responsiveness according to in-session engagements
Newsfeed: diversity re-ranker to ensure calibrated mixture of topics, text, video, and links

He received his PhD from Tsinghua University, advised by Jianyong Wang, was a visiting scholar at UIUC, advised by Jiawei Han. He got bachelor degree in software engineering from Beihang University.

Publications

TorchTitan: One-stop PyTorch Native Solution for Production Ready LLM Pre-training
Wanchao Liang, Tianyu Liu, Less Wright, Will Constable, Andrew Gu, Chien-Chin Huang, Iris Zhang, Wei Feng, Howard Huang, Junjie Wang, Sanket Purandare, Gokul Nadathur, Stratos Idreos
ICLR 2025 · 2025
PDF
TORCHSIM: High Fidelity Runtime and Memory Estimation for Distributed Training
Sanket Purandare, Emma Yang, Andrew Zhao, Qitong Wang, Wei Feng, Alban Desmaison, Andrew Gu, Tianyu Liu, Less Wright, Gokul Nadathur, Stratos Idreos
ES-FoMo III: 3rd Workshop on Efficient Systems for Foundation Models 2025 · 2025
PDF
Dynamically Pruned Message Passing Networks for Large-Scale Knowledge Graph Reasoning
X Xu, W Feng, Y Jiang, X Xie, Z Sun, ZH Deng
ICLR 2020 · 2020
PDF
Neural Consciousness Flow
X Xu, W Feng, Z Sun, ZH Deng
Preprint · 2019
PDF
Backprop-Q: Generalized Backpropagation for Stochastic Computation Graphs
X Xu, S Zu, Y Zhang, H Zhou, W Feng
NeurIPS 2018 Deep Reinforcement Learning Workshop · 2018
PDF
Modeling Attention Flow on Graphs
X Xu, S Zu, C Gao, Y Zhang, W Feng
NeurIPS 2018 Relational Representation Learning Workshop · 2018
PDF
StreamCube: Hierarchical Spatio-temporal Hashtag Clustering for Event Exploration over the Twitter Stream
Wei Feng, Jiawei Han, Jianyong Wang, Charu Aggarwal, Jianbin Huang
ICDE 2015 · 2015
PDF
We Can Learn Your #Hashtags: Connecting Tweets to Explicit Topics
Wei Feng, Jianyong Wang, Wei Zhang
ICDE 2014 · 2014
Paper
Combining Latent Factor Model with Location Features for Event-based Group Recommendation
Wei Zhang, Jianyong Wang, Wei Feng
KDD 2013 · 2013
Paper
Integrating Semantic Relatedness and Words' Intrinsic Features for Keyword Extraction
Wei Zhang, Wei Feng, Jianyong Wang
IJCAI 2013 · 2013
PDF
Learning to Annotate Tweets with Crowd Wisdom
Wei Feng, Jianyong Wang
WWW 2013 (Poster) · 2013
Paper
Retweet or Not? Personalized Tweet Re-ranking
Wei Feng, Jianyong Wang
WSDM 2013 · 2013
Paper
Incorporating Heterogeneous Information for Personalized Tag Recommendation in Social Tagging Systems
Wei Feng, Jianyong Wang
KDD 2012 · 2012
Paper
HeterRank: Addressing Information Heterogeneity for Personalized Recommendation in Social Tagging Systems
Wei Feng, Jianyong Wang
WWW 2012 (Poster) · 2012
Paper