Portrait
Wonjung Kim
M.S. student
AI Computing, KAIST
About Me

I am an M.S. student at VLlab, KAIST, supervised by Prof. Seunghoon Hong, specializing in computer vision and generative modeling. I specialize in high-fidelity generative modeling for image and text domains, with a primary focus on few-step synthesis. My work extends to enhancing the efficiency of both training and inference processes for generative models.

Education
  • Korea Advanced Institute of Science and Technology (KAIST)
    Korea Advanced Institute of Science and Technology (KAIST)
    M.S. in AI Computing
    Sep. 2026 - Aug. 2028 (expected)
  • Korea Advanced Institute of Science and Technology (KAIST)
    Korea Advanced Institute of Science and Technology (KAIST)
    B.S. in School of Computing
    Mar. 2020 - Aug. 2026
Experience
  • Vision and Language Laboratory (VLlab)
    Vision and Language Laboratory (VLlab)
    Research Intern
    Jan. 2025 - present
News
2026
IMDM has been accepted at ICML 2026
May 01
2025
ReDi has been accepted at NeurIPS 2025
Sep 19
Publications (view all )
Self-conditioned Flow Map Language Models via Fixed-point Flows
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 (* equal contribution, equal advising)

Preprint.

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 (* equal contribution, equal advising)

Preprint.

Infinite Mask Diffusion for Few-Step Distillation
Infinite Mask Diffusion for Few-Step Distillation

Jaehoon Yoo*, Wonjung Kim*, Chanhyuk Lee, Seunghoon Hong (* equal contribution)

ICML 2026

Infinite Mask Diffusion for Few-Step Distillation

Jaehoon Yoo*, Wonjung Kim*, Chanhyuk Lee, Seunghoon Hong (* equal contribution)

ICML 2026

ReDi: Rectified Discrete Flow
ReDi: Rectified Discrete Flow

Jaehoon Yoo, Wonjung Kim, Seunghoon Hong

NeurIPS 2025

ReDi: Rectified Discrete Flow

Jaehoon Yoo, Wonjung Kim, Seunghoon Hong

NeurIPS 2025

All publications