Jinwoo Kim

jinwoo-kim [at] kaist.ac.kr

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 beyond their training data so that they can be used to solve challenging problems, such as those in scientific domains. I am studying this problem from the complementary perspectives of invariance and equivariance, parallel neural processing, and Solomonoff induction. I often use tools from theories of graphs, manifolds, (semi)groups and categories, as well as Markov processes such as random walks, diffusions, and flows.

Recently, I am focused on flow language models, latent reasoning, synthetic data, and vision pretraining.

Experience

KAIST - Expert Research Personnel (what?), 03/2026 – 02/2027
New York University - Visiting Scholar, 11/2025 – 01/2026
LG AI Research - Research Intern, 2022

Ph.D. Thesis

Architecture-Agnostic Invariances for Deep Learning

Jinwoo Kim

Korea Advanced Institute of Science & Technology, School of Computing, 2026

KAIST CoE Best Dissertation Award

pdf

Selected Papers

Thinking with Looped Flows

Ayhan Suleymanzade, Chanhyuk Lee, Floor Eijkelboom, Nicholas M. Boffi, İsmail İlkan Ceylan^†, Jinwoo Kim^†

arXiv 2026

paper

MUX: Continuous Reasoning via Multiplexed Tokens

Ayhan Suleymanzade, Halil Alperen Gozeten, Michael Bronstein, İsmail İlkan Ceylan^†, Jinwoo Kim^†

NeurIPS 2026 Spotlight (0.95%)

paper code

Flow Map Language Models: One-step Language Modeling via Continuous Denoising

Chanhyuk Lee, Jaehoon Yoo, Manan Agarwal, Sheel Shah, Jerry Huang, Aditi Raghunathan, Seunghoon Hong, Nicholas M. Boffi^†, Jinwoo Kim^†

NeurIPS 2026 / COLM NonAR-LM Workshop 2026 Spotlight

paper code blog

Self-conditioned Flow Map Language Models via Fixed-point Flows

Jaehoon Yoo*, Wonjung Kim*, Floor Eijkelboom, Chanhyuk Lee, Nicholas M. Boffi, Seunghoon Hong^†, Jinwoo Kim^†

COLM NonAR-LM Workshop 2026 Oral

paper code

Posterior Refinement: Fast Language Generation via Any-Order Flow Maps

Manan Agarwal*, Sheel Shah*, Chanhyuk Lee, Jaehoon Yoo, Jerry Huang, Seunghoon Hong, Aditi Raghunathan, Jinwoo Kim^†, Nicholas M. Boffi^†

COLM NonAR-LM Workshop 2026 Spotlight

paper code

What are the Right Symmetries for Formal Theorem Proving?

Krzysztof Olejniczak, Radoslav Dimitrov, Xingyue Huang, Bernardo Cuenca Grau, Jinwoo Kim^†, İsmail İlkan Ceylan^†

ICML AI4Math Workshop 2026

paper code

Inverting Data Transformations via Diffusion Sampling

Jinwoo Kim*, Sékou-Oumar Kaba*, Jiyun Park, Seunghoon Hong^†, Siamak Ravanbakhsh^†

ICML 2026

paper code

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

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

Full list of publications

Honors

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

Invited Talks

Flow language models
@ Discrete Diffusion Reading Group (04/2026), SNU (07/2026), POSTECH (10/2026)
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 for few-shot dense vision
@ 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 Workshop (01/2022), NeurIPS Social: ML in Korea (12/2021)
Variational autoencoders for sets
@ Naver AI (09/2021), KCCV (09/2021)
Retinal waves and prenatal wiring of primary visual cortex
@ Society for Neuroscience, 10/2019

Academic Services

Area Chair: ICLR 2027
Workshop Area Chair: ICML 2026 GFM
Conference Reviewer: NeurIPS, ICLR, ICML, CVPR, ICCV, AISTATS, LoG, TAG-DS, IJCNN, ACCV
Journal Reviewer: TMLR, Neural Networks, IJCV

Teaching

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

Last updated: September 2026