Statistical Foundations of Latent Structure and Representation Learning
November 4–5, 2026
Department of Statistics, Columbia University New York City
About
Many modern problems in statistics and machine learning depend on uncovering latent structure and learning informative representations from complex, high-dimensional data. These problems raise fundamental questions about identifiability, statistical efficiency, uncertainty quantification, computation, causality, and interpretability, with connections spanning latent variable models, high-dimensional statistics, networks, and modern representation learning.
This two-day workshop brings together researchers from statistics, machine learning, data science, and related fields to discuss recent advances and explore emerging connections across these areas, with an emphasis on the statistical foundations of latent structure and representation learning.
Registration
Registration is free and open to interested participants. Advance registration is requested to help us plan seating and refreshments. Please register by if possible. Late registration may remain available subject to space.
Register for the Workshop (opens in a new tab)
PhD students, postdoctoral researchers, and junior faculty interested in presenting a poster should use the separate poster submission form below.
Speakers
Speakers are listed alphabetically by last name.
Florentina Bunea
Cornell University
Tianxi Cai
Harvard University
Yuxin Chen
University of Pennsylvania
David Dunson
Duke University
Jianqing Fan
Princeton University
Biwei Huang
UC San Diego
Jiashun Jin
Carnegie Mellon University
George Michailidis
UCLA
Carey Priebe
Johns Hopkins University
Annie Qu
UC Santa Barbara
Veronika Rockova
University of Chicago
Ali Shojaie
University of Washington
Gongjun Xu
University of Michigan
Bin Yu
UC Berkeley
Kun Zhang
Carnegie Mellon University
Hongtu Zhu
UNC Chapel Hill
Ji Zhu
University of Michigan
Hui Zou
Johns Hopkins University
Early-Career Researcher Poster Session
An early-career researcher poster session is tentatively planned for approximately 4:30–6:00 p.m. on November 4. Early-career researchers (including PhD students, postdoctoral researchers, and junior faculty) are invited to submit work broadly related to latent structure, representation learning, high-dimensional statistics, causal learning, statistical machine learning, and related areas. The session will provide early-career researchers with an opportunity to present their work and interact with workshop speakers and participants.
- Priority submission deadline
- Expected notification
- Around
- Submission
- Poster title and a short abstract of 250 words or fewer
A limited number of travel-support awards of up to $500 per presenter will be available to help nonlocal PhD students and postdoctoral presenters offset workshop-related transportation and lodging expenses.
Program
Detailed program forthcoming.
The workshop will take place November 4–5. The early-career researcher poster session is tentatively scheduled for approximately 4:30–6:00 p.m. on November 4, and the workshop is expected to conclude on the afternoon of November 5. The exact schedule is being finalized; talk titles and the detailed program will be posted as they become available.
Location
Columbia University, New York City.
The exact venue and logistical details will be posted here when finalized.
Organizer
Yuqi Gu
Department of Statistics, Columbia University
Questions about the workshop: yuqi.gu@columbia.edu