Research Interests
I am broadly interested in mathematical and algorithmic foundations of machine learning for distributions and dynamical systems. My work draws on optimal transport, including primal and low-rank methods; variational optimization in Wasserstein space, including the JKO Scheme, Wasserstein gradient flows, and Fokker-Planck dynamics; and modern generative modeling approaches based on score-matching and diffusion models.
I apply these mathematical and machine-learning tools in AI for science, particularly to high-throughput single-cell data measured across time and space. I develop methods based on optimal transport, geometric deep learning, differential geometry, and related mathematical tools to infer correspondences between snapshot observations and reconstruct spatiotemporal and multiomic dynamics.
Blog Notes & News
Blog posts on optimal transport, machine learning, information geometry, and biology.
News
Riemannian Metric Learning for Alignment of Spatial Multiomics received the ISMB Best Student Paper Award!
Publications
Transport Clustering: Solving Low-Rank Optimal Transport via Clustering
ICML 2026 | Poster (PDF)
Riemannian Metric Learning for Alignment of Spatial Multiomics
ISMB 2026 (Best Student Paper, Ian Lawson Van Toch Memorial Award); Bioinformatics 2026 (Systems biology and networks)
Multimodal spatial alignment and morphology mapping with MOSAICField
RECOMB 2026
Implicit Bias of the JKO Scheme
Hierarchical Refinement: Optimal Transport to Infinity and Beyond
ICML 2025 (Oral)
Learning Latent Trajectories in Developmental Time Series with Hidden-Markov Optimal Transport
RECOMB 2025
Low-Rank Optimal Transport through Factor Relaxation with Latent Coupling
NeurIPS 2024 | Slides (Keynote)
DeST-OT: Alignment of Spatiotemporal Transcriptomics Data
Cell Systems 2025 · RECOMB 2024
System Identification for Continuous-time Linear Dynamical Systems
Abstracts
Pooled RNA-IP approach to investigate variant effects on RBP binding and splicing
ASHG 2022 (Poster)
Pan-cancer analysis of sex differences and their associations with ancestry and genomic biomarkers in a large comprehensive genomic profiling dataset
AACR 2020
Pan-chromosome analysis does not reveal parental or ancestral bias in chromosome loss of heterozygosity
AACR 2020 (Accepted but not presented)
Talks
A selection of recent talks. Slides are linked when available.
Riemannian Metric Learning for Alignment of Spatial Multiomics
Optimal Transport Modeling of Cellular Differentiation: From Low-Rank Structure to Temporal Dynamics
Hierarchical Refinement: Optimal Transport to Infinity and Beyond
Hidden-Markov Optimal Transport for Developmental Time Series
Teaching and Service
Teaching Assistant: Information Theory (COS 585, Princeton 2026), Organic Chemistry (CHEM UN2443, Columbia 2021)
Instructor: Introduction to Computer Science (King Summer Institute 2023)
Reviewer: TMLR 2026, NeurIPS 2026, Transactions on Pattern Analysis and Machine Intelligence 2026, ICML 2026, NeurIPS 2025, ISMB/ECCB 2025, RECOMB 2025, RECOMB 2024.
Graduate mentorship: Princeton Pre-Application PhD Program, First Year Mentorship Program.
Relevant Coursework
Partial Differential Equations (MAT522), Information Theory (COS 585), Differential Geometry (MAT 550), Theoretical Machine Learning (COS 511), Dynamical Systems (APC 571), Deep Learning Theory (ORF 543), Statistical Mechanics (CHEM3079), Machine Learning and Pattern Recognition (ECE 535), Advanced Algorithm Design (COS 521), Ordinary Differential Equations (MATHUN2030), Measure-Theoretic Probability (MAT 385), Stochastic Calculus (ORF 527; AUD), Advanced Organic Chemistry (CHEM GU4147), Biochemistry (BCHM 4501).
Other
Outside of my academic interests, I am an avid hiker and an amateur birder, botanist, and photographer.