Statistical Foundations of Generative Modeling | Poster Session

Presenter: Vijay Kumar (Habib University, Brown University)
Collaborator(s): Joon Jin Song(Baylor University), Yuzi Zhong(The Ohio State University), Rebbeca Kurtz-Garcia(Smith College), Nan Lin(WashU St. Louis) and Zainab Mohammad(Oberlin College)
Title: Bayesian Generative Calibration of Low-Cost Air Quality Sensors Data

Presenter: Shucheng Liu (University of North Carolina Chapel Hill)
Collaborator(s): Changchun Shi, Kai Zhang, Hongtu Zhu
Title: FAST-Brain: A Flow-Aligned Spatio-Temporal Surrogate Brain Model

Presenter: Pengfei Lyu (Duke University)
Collaborator(s):
Title: Bias-Corrected Data Synthesis for Imbalanced Learning

Presenter: Dushyant Mehra (University of Texas Southwestern Medical Center)
Collaborator(s): Harish Srinivasan, Bassel Dawod, Jian Zhou, Kevin Dean, Payal Kapur, Todd Aguilera
Title: Reward Guided H&E-to-mIF Translation using via Multimodal Flow Maps with Compact Representation Autoencoders for Treatment Response Prediction

Presenter: Huy Nguyen (University of Texas, Austin)
Collaborator(s): Thong Doan, Quang Pham, Nghi Bui, Nhat Ho, Alessandro Rinaldo
Title: On DeepSeekMoE: Statistical Benefits of Shared Experts and Normalized Sigmoid Gating

Presenter: Fred Peng (Duke University)
Collaborator(s): Avishek Joey Bose (Imperial College London); Anru R. Zhang (Duke University); Alexander Tong (AITHYRA)
Title: Coupling Models for One-Step Discrete Generation

Presenter: Zihua She (Purdue University)
Collaborator(s): Jianxi Su, Xiao Wang
Title: Coreset-Based Joint Source-Coupling Design for Flow Matching

Presenter: Borui Tang (University of North Carolina Chapel Hill)
Collaborator(s): Yue Yang, Ting Li, Zehui Sun, Tengfei Li, Hongtu Zhu
Title: Integrative Analysis of Plasma Proteomics and Magnetic Resonance Imaging for Alzheimer's Disease and Cognitive Function

Presenter: Amanda Wang (Cornell University)
Collaborator(s): Tudor Manole, Florentina Bunea, John Thickstun
Title: Fidelity-Diversity Metrics for Text Datasets

Presenter: xiaoyi wen (University of California, San Francisco (UCSF))
Collaborator(s): Fei Jiang
Title: IWCAE: Importance-Weighted Conditional Autoencoders for Domain-Adaptive Generative Modeling

Presenter: Sichen Yang (Johns Hopkins University)
Collaborator(s): Mauro Maggioni
Title: Multi-level Meta-Reinforcement Learning with Skill-Based Curriculum

Presenter: Cheng Yu (University of Chicago)
Collaborator(s):
Title: Efficient Design and Inference for Long-Term Holdout Experiments with Predictive Distributions

Presenter: Yikun Zhang (University of Chicago and SkAI Institute)
Collaborator(s): Steven Wilkins-Reeves, Wesley Lee, Aude Hofleitner
Title: Transfer Learning Through Conditional Quantile Matching