Yangsibo Huang

Welcome! I am a fourth-year PhD student at Princeton University. I am very fortunate to be advised by Prof. Kai Li. I am also working closely with Prof. Sanjeev Arora. My research interests lie in the interface of Systems and Machine Learning, especially the privacy and security topics. I am also interested in machine learning for healthcare(ML4H).

Previously, I got my bachelor’s degree in Computer Science at Zhejiang University. I spent a great semester in my senior year at Harvard Medical School and Massachusetts General Hospital under the supervision of Dr. Quanzheng Li. I also spent a wonderful summer at Prof. Ke Sheng’s lab at David Geffen School of Medicine at UCLA.

News

[09/2022] Our paper entitled “Recovering Private Text in Federated Learning of Language Models” has been accepted by NeurIPS 2022. We proposed an attack that recovers private text in Federated Learning of language models for a batch size up to 128 (Check out the code here)!
[08/2022] Our proposal on data auditing for machine learning models won Meta’s 2022 Privacy-Enhancing Technologies request for proposals.
[01/2022] Proposal about privacy-preserving FL funded by Princeton University’s Schmidt DataX Fund.
[11/2021] We’ve released the GradAttack library! Interested in gradient inversion attacks and defenses in Federated Learning? Come and play with it!
[09/2021] Our paper entitled “Evaluating Gradient Inversion Attacks and Defenses in Federated Learning” has been accepted by NeurIPS 2021 as an Oral (<1% acceptance rate). We evaluated recent gradient inversion attacks and defenses in federated learning (and summarized some best practices).

Selected Publications and Manuscripts

2022

Recovering Private Text in Federated Learning of Language Models

Samyak, Gupta, Yangsibo, Huang, Zexuan, Zhong, Tianyu, Gao, Kai, Li, and Danqi, Chen

In Advances in Neural Information Processing Systems (NeurIPS) 2022

Paper Code

2021

Evaluating Gradient Inversion Attacks and Defenses in Federated Learning

Yangsibo, Huang, Samyak, Gupta, Zhao, Song, Kai, Li, and Sanjeev, Arora

In Advances in Neural Information Processing Systems (NeurIPS) 2021 (Oral)

Paper Code Poster
EMA: Auditing Data Removal from Trained Models

Yangsibo, Huang, Xiaoxiao, Li, and Kai, Li

In International Conference on Medical Image Computing and Computer Assisted Intervention (MICCAI) 2021

Paper Code
DeepMC: a deep learning method for efficient Monte Carlo beamlet dose calculation by predictive denoising in magnetic resonance-guided radiotherapy

Ryan, Neph, Qihui, Lyu, Yangsibo, Huang, You Ming, Yang, and Ke, Sheng

Physics in Medicine & Biology 2021

Paper

2020

TextHide: Tackling Data Privacy in Language Understanding Tasks

Yangsibo, Huang, Zhao, Song, Danqi, Chen, Kai, Li, and Sanjeev, Arora

In The Conference on Empirical Methods in Natural Language Processing (Findings of EMNLP) 2020

Paper Code
InstaHide: Instance-hiding Schemes for Private Distributed Learning

Yangsibo, Huang, Zhao, Song, Kai, Li, and Sanjeev, Arora

In Internation Conference on Machine Learning (ICML) 2020

Paper Code

2019

Deep Q learning Driven CT Pancreas Segmentation with Geometry-aware U-Net

Yunze, Man, Yangsibo, Huang, Junyi, Feng, Xi, Li, and Fei, Wu

IEEE Transactions on Medical Imaging 2019

Paper

Reviewer & committee

Program Committee member for workshop on Interpretable Machine Learning in Healthcare (co-located with ICML 2021, 2022)
Program Committee member for workshop on Computer Vision for Automated Medical Diagnosis (co-located with ICCV 2021)
Reviewer for ICML (2022), NeurIPS (2021, 2022)
Reviewer for IEEE Transactions on Medical Imaging (TMI)

Teaching Assistant

(20-21 S’) ECE 382: Probabilistic Systems and Information Processing

Contact me

You are welcome to reach out to me by email or via WeChat (ID: hazelsuko07).