I'm a PhD candidate at Stanford CS. I go by Chen, the second segment of my first name.
My research revolves around machine learning, deep learning, and NLP.
My current focus is large language model training, data collection/curation, and human evaluation.
Here are some specific topics of interest:
Data and Evaluation: Recent improvements in AI systems have been partially driven by training with high-quality human data.
I am interested in (alternative) scalable data collection paradigms that leverage novel incentive structures.
As models become more capable, their evaluation becomes increasingly challenging.
I am also interested in developing new evaluations that measure not only capability but also usability.
Learning From Human Feedback: Human feedback has become a primary driver for recent successes in AI, such as ChatGPT.
However, collecting and training on such data can be costly and cumbersome. Some of the questions I'm recently interested in
are: How can we efficiently elicit high-quality feedback? How can we augment the feedback data when it comes in limited quantities?
How should we aggregate this feedback signal without marginalizing the minority voices and views?
Red Teaming and Auditing: Despite the rapid progress in capability research, machine learning models still exhibit systematic flaws. I am interested in developing automated tools to assist humans in identifying and rectifying these flaws.
Memorization and Privacy: Large models can memorize training data. This not only poses privacy risks but also raises emergent sociotechnical questions (e.g., on copyright and intellectual property). I am interested in understanding this memorization phenomenon and in developing tools to mitigate its undesirable consequences. Here is an outdated statement I
wrote on privacy and security in machine learning. Some of my research has seen growing adoption in industry.
Selected and Recent Research (full list see google
AlpacaFarm: A Simulation Framework for Methods that Learn from Human Feedback
Yann Dubois*, Xuechen Li*, Rohan Taori*, Tianyi Zhang*, Ishaan Gulrajani, Jimmy Ba, Carlos Guestrin, Percy Liang, and Tatsunori B. Hashimoto
Alpaca: A Strong, Replicable Instruction-Following Model
Rohan Taori*, Ishaan Gulrajani*, Tianyi Zhang*, Yann Dubois*, Xuechen Li*, Carlos Guestrin, Percy Liang, and Tatsunori B. Hashimoto
Foundation Models and Fair Use
Peter Henderson*, Xuechen Li*, Dan Jurafsky, Tatsunori Hashimoto, Mark A. Lemley, Percy Liang
Exploring the Limits of Differentially Private Deep Learning with Group-wise Clipping
Jiyan He*, Xuechen Li*, Da Yu*, Huishuai Zhang, Janardhan Kulkarni, Yin Tat Lee, Arturs Backurs, Nenghai Yu, Jiang Bian
International Conference on Learning Representations, 2023
Large Language Models Can Be Strong Differentially Private Learners
Xuechen Li, Florian Tramer, Percy Liang, Tatsunori Hashimoto
International Conference on Learning Representations, 2022
NeurIPS Privacy in Machine Learning Workshop, 2021
When Does Preconditioning Help or Hurt Generalization?
Shun-ichi Amari, Jimmy Ba, Roger Grosse, Xuechen Li, Atsushi Nitanda, Taiji Suzuki, Denny Wu, Ji Xu
International Conference on Learning Representations, 2021
12th OPT Workshop on Optimization for ML
[Best student paper]
Scalable Gradients for Stochastic Differential Equations
Xuechen Li, Ting-Kam Leonard Wong, Ricky T. Q. Chen, David Duvenaud
International Conference on Artificial Intelligence and Statistics, 2020
2nd Symposium on Advances in Approximate Bayesian Inference
Stochastic Runge-Kutta Accelerates Langevin Monte Carlo and Beyond
Xuechen Li, Denny Wu, Lester Mackey, Murat A. Erdogdu
Advances in Neural Information Processing Systems, 2019