Yunbum Kook
Office: 2210 KACB
Email: yb.{last name} at gatech dot edu

I am a fourth-year PhD student in the School of Computer Science at Georgia Institute of Technology, advised by Prof. Santosh Vempala. I received my B.S. in Mathematical Sciences from KAIST. I am interested in sampling, diffusion models, and theoretical aspects of machine learning.

Scholar  

Publications

Covariance estimation using Markov chain Monte Carlo
with Matthew S. Zhang
arXiv  

Rényi-infinity constrained sampling with d^3 membership queries
with Matthew S. Zhang
arXiv   SODA 2025   Slide  

In-and-Out: Algorithmic diffusion for sampling convex bodies
with Santosh S. Vempala, Matthew S. Zhang
arXiv   [NeurIPS 2024 (Spotlight)]   Poster   Slide  

Sampling from the mean-field stationary distribution
with Sinho Chewi, Murat A. Erdogdu, Mufan Bill Li, Matthew S. Zhang
arXiv   COLT 2024   Poster   Slide  

Gaussian cooling and Dikin walks: The interior-point method for logconcave sampling
with Santosh S. Vempala
arXiv   COLT 2024   Poster   Slide  

Understanding Adam optimizer via online learning of updates: Adam is FTRL in disguise
Kwangjun Ahn, Zhiyu Zhang, Yunbum Kook, Yan Dai
arXiv   ICML 2024   Video (15)  

Condition-number-independent convergence rate of Riemannian Hamiltonian Monte Carlo with numerical integrators
with Yin Tat Lee, Ruoqi Shen, Santosh S. Vempala
arXiv   COLT 2023   Poster   Slide  

Sampling with Riemannian Hamiltonian Monte Carlo in a constrained space
with Yin Tat Lee, Ruoqi Shen, Santosh S. Vempala
arXiv   NeurIPS 2022   Poster   Slide   Code  

Vertex sparsification for edge connectivity
with Parinya Chalermsook, Syamantak Das, Bundit Laekhanukit, Yang P. Liu, Richard Peng, Mark Sellke, Daniel Vaz
arXiv  SODA 2021  Slide   Video(11)   Video(25)   Video(60)  

Evolution of real-world hypergraphs: Patterns and models without oracles (long)
Yunbum Kook, Jihoon Ko, Kijung Shin
arXiv   ICDM 2020   [Knowledge and Information Systems]   Slide   Video(20)   Code  

Incremental lossless graph summarization
Jihoon Ko*, Yunbum Kook*, Kijung Shin
arXiv   KDD 2020   Slide   Video(20)   Code  

Template credits