jinwoo-kim [at] kaist.ac.kr
CV / Scholar / GitHub / X / LinkedIn
I am a technical research personnel at KAIST. I was previously a visiting scholar at NYU Global AI Frontier Lab working with Kyunghyun Cho and Rajesh Ranganath. I received my Ph.D. from KAIST advised by Seunghoon Hong. My name 진우 眞友 is pronounced [jeen-oo].
I am interested in making deep learning models generalize better out of their training data so that they can be used to solve challenging problems, such as those in scientific domains. I have been studying this problem primarily from the viewpoint of invariances, focusing on all-purpose deep neural networks that can reliably reason upon novel transformed inputs. I often use tools from theories of graphs, (semi)groups and manifolds, and Markov processes such as random walks, diffusions and flows.
KAIST - Expert Research Personnel (what?), 03/2026 – 02/2027
New York University - Visiting Scholar, 11/2025 – 01/2026
LG AI Research - Research Intern, 2022
Architecture-Agnostic Invariances for Deep Learning
Jinwoo Kim
Ph.D. Dissertation, 2026. KAIST CoE Best Dissertation Award
[pdf]
Flock: A Knowledge Graph Foundation Model via Learning on Random Walks
Jinwoo Kim*, Xingyue Huang*, Krzysztof Olejniczak, Kyungbin Min, Michael Bronstein, Seunghoon Hong, İsmail
İlkan
Ceylan
ICLR, 2026
[paper] [code]
Sequence Modeling with Spectral Mean Flows
Jinwoo Kim, Max Beier, Petar Bevanda, Nayun Kim, Seunghoon Hong
NeurIPS, 2025
[paper] [code]
Revisiting Random Walks for Learning on Graphs
Jinwoo Kim, Olga Zaghen*, Ayhan Suleymanzade*, Youngmin Ryou, Seunghoon Hong
ICLR, 2025. Spotlight (3%)
[paper] [code]
Learning Probabilistic Symmetrization for Architecture Agnostic Equivariance
Jinwoo Kim, Tien Dat Nguyen, Ayhan Suleymanzade, Hyeokjun An, Seunghoon Hong
NeurIPS, 2023. Spotlight (3%)
[paper] [code]
Universal Few-shot Learning of Dense Prediction Tasks with Visual Token Matching
Donggyun Kim, Jinwoo Kim, Seongwoong Cho, Chong Luo, Seunghoon Hong
ICLR, 2023. Outstanding Paper Award (0.08%)
[paper] [code]
Pure Transformers are Powerful Graph Learners
Jinwoo Kim, Tien Dat Nguyen, Seonwoo Min, Sungjun Cho, Moontae Lee, Honglak Lee†, Seunghoon
Hong†
NeurIPS, 2022
[paper] [code]
Best Ph.D. Dissertation Award, KAIST College of Engineering, 2026
Outstanding Researcher Award, KAIST-Mila Prefrontal AI Research Center, 2024
ELLIS Mobility Grant, ICML 2024 GRaM Workshop
Outstanding Paper Award, ICLR 2023 (as coauthor)
Silver Prize, Samsung Humantech Paper Award, 2023 (as coauthor)
Qualcomm Innovation Fellowship Korea, 2022
Excellence Award, KAIST Undergraduate Research Program, 2022 (as mentor)
Kwanjeong Education Foundation Scholarship, 2022–2023
KAIST Engineering Innovator Award, 2020
National Science & Technology Scholarship, 2018–2020
KAIST Alumni Fellowship, 2017–2020
KAIST Presidential Fellowship, 2016–2020
KAIST Dean's List, Spring 2016 / Fall 2016 / Spring 2018
Sequence Modeling with Spectral Mean Flows @ BGU, 12/2025
Architecture-Agnostic Invariances for Deep Learning @ Mila (07/2025, 12/2024), KAIST AI899 (05/2025), KAIST-Mila
Prefrontal AI Research Center (11/2024), SKKU (08/2024), POSTECH (11/2023)
Visual Token Matching @ KAIST-Samsung Electronics DS
Division Exchange Meetup, 08/2023
Pure Transformers are Powerful Graph Learners @ Microsoft USA (01/2023), NeurIPS 2022 at KAIST (11/2022), LoGaG
(08/2022)
Higher-order Transformers for Sets, Graphs, and Hypergraphs @ Qualcomm Korea (01/2023), KAIST AI
Workshop (01/2022), NeurIPS Social: ML in Korea (12/2021)
SetVAE @ Naver AI (09/2021), KCCV (09/2021)
Retinal Waves and Prenatal Wiring of Primary Visual Cortex @ SfN, 10/2019
Conference Reviewer: NeurIPS, ICLR, ICML, CVPR, ICCV, AISTATS, LoG, TAG-DS, IJCNN, ACCV
Journal Reviewer: TMLR, Neural Networks, IJCV
TA, Computer Vision (CS576), Spring 2022/2023
TA, Introduction to Deep Learning (CS492I/CS371), Fall 2021/2022/2023
TA, Samsung Research AI Expert Program, Summer 2021/2022/2023
TA, Undergraduate Research Program (URP), Spring 2022/2024
I also enjoy making music. Some of my compositions: song 1 song 2
Last updated: February 2026