Lechao Xiao

Researcher, Google DeepMind

Joined

October 2018

Names

Lechao Xiao (Preferred)

Emails

****@gmail.com (Confirmed)

****@google.com (Confirmed)

****@gmail.com (Confirmed)

Personal Links

Homepage

Google Scholar

DBLP

Career & Education History

Researcher

Google DeepMind (google.com)

2023 – Present

Research Scientist

Google Research, Brain Team (google.com)

2017 – 2023

Google Brain Resident

Google Research, Brain Team (google.com)

2017 – 2019

Postdoc

University of Pennsylvania (upenn.edu)

2014 – 2017

PhD student

University of Illinois, Urbana-Champaign (uiuc.edu)

2009 – 2014

Advisors, Relations & Conflicts

No relations added

Expertise

Mathematics of deep learning/ machine learning

Present

deep learning theory

2017 – Present

mathematics

2009 – Present

Publications

Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks
Shikai Qiu, Lechao Xiao, Andrew Gordon Wilson, Jeffrey Pennington, Atish Agarwala
- ICML 2025 oral
- Readers: Everyone
Scaling Collapse Reveals Universal Dynamics in Compute-Optimally Trained Neural Networks
Shikai Qiu, Atish Agarwala, Jeffrey Pennington, Lechao Xiao
- NeurIPS 2024 Workshop
- Readers: Everyone
4+3 Phases of Compute-Optimal Neural Scaling Laws
Elliot Paquette, Courtney Paquette, Lechao Xiao, Jeffrey Pennington
- NeurIPS 2024 spotlight
- Readers: Everyone
Training Language Models on the Knowledge Graph: Insights on Hallucinations and Their Detectability
Jiri Hron, Laura A Culp, Gamaleldin Fathy Elsayed, Rosanne Liu, Jasper Snoek, Simon Kornblith, Alex Rizkowsky, Isabelle Simpson, Jascha Sohl-Dickstein, Noah Fiedel, Aaron T Parisi, Alexander A Alemi, Azade Nova, Ben Adlam, Bernd Bohnet, Gaurav Mishra, Hanie Sedghi, Izzeddin Gur, Jaehoon Lee, John D Co-Reyes et al. (11 additional authors not shown)
- COLM
- Readers: Everyone
Scaling Exponents Across Parameterizations and Optimizers
Katie E Everett, Lechao Xiao, Mitchell Wortsman, Alexander A Alemi, Roman Novak, Peter J Liu, Izzeddin Gur, Jascha Sohl-Dickstein, Leslie Pack Kaelbling, Jaehoon Lee, Jeffrey Pennington
- ICML 2024 Poster
- Readers: Everyone
Beyond Human Data: Scaling Self-Training for Problem-Solving with Language Models
Avi Singh, John D Co-Reyes, Rishabh Agarwal, Ankesh Anand, Piyush Patil, Xavier Garcia, Peter J Liu, James Harrison, Jaehoon Lee, Kelvin Xu, Aaron T Parisi, Abhishek Kumar, Alexander A Alemi, Alex Rizkowsky, Azade Nova, Ben Adlam, Bernd Bohnet, Gamaleldin Fathy Elsayed, Hanie Sedghi, Igor Mordatch et al. (20 additional authors not shown)
- Accepted by TMLR
- Readers: Everyone
Small-scale proxies for large-scale Transformer training instabilities
Mitchell Wortsman, Peter J Liu, Lechao Xiao, Katie E Everett, Alexander A Alemi, Ben Adlam, John D Co-Reyes, Izzeddin Gur, Abhishek Kumar, Roman Novak, Jeffrey Pennington, Jascha Sohl-Dickstein, Kelvin Xu, Jaehoon Lee, Justin Gilmer, Simon Kornblith
- ICLR 2024 oral
- Readers: Everyone
Fast Neural Kernel Embeddings for General Activations
Insu Han, Amir Zandieh, Jaehoon Lee, Roman Novak, Lechao Xiao, Amin Karbasi
- Published: 31 Oct 2022, Last Modified: 08 Feb 2026
- NeurIPS 2022 Accept
- Readers: Everyone
Precise Learning Curves and Higher-Order Scalings for Dot-product Kernel Regression
Lechao Xiao, Hong Hu, Theodor Misiakiewicz, Yue Lu, Jeffrey Pennington
- Published: 31 Oct 2022, Last Modified: 08 Feb 2026
- NeurIPS 2022 Accept
- Readers: Everyone
Eigenspace Restructuring: a Principle of Space and Frequency in Neural Networks
Lechao Xiao
- Published: 28 Jan 2022, Last Modified: 12 Oct 2025
- ICLR 2022 Submitted
- Readers: Everyone

View all 15 publications

Co-Authors

View all 59 co-authors