Skip to content
View taehoon-yoon's full-sized avatar
⭐
⭐
21 followers · 11 following
  • KAIST
  • Seoul. Korea
  • 20:31 (UTC +09:00)

Achievements

Achievement: StarstruckAchievement: Pull SharkAchievement: Quickdraw

Achievements

Achievement: StarstruckAchievement: Pull SharkAchievement: Quickdraw

Highlights

  • Pro

Block or report taehoon-yoon

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
taehoon-yoon/README.md

TAEHOON YOON

ACADEMIC INTERESTS

Generative Modeling, Probabilistic Inference:

Diffusion and Flow-Matching Models, Controllable Generation, Reward Alignment, Inverse Problems, Particle-Based Sampling Methods

EDUCATION

  • Korea Advanced Institute of Science and Technology (KAIST)    [Mar. 2024 - Present]

    M.S. in Artificial Intelligence (Graduate School of Artificial Intelligence)

  • Sogang University  [Mar. 2017 - Feb. 2024]

    B.S. in Physics & Computer Science and Engineering (Double Major),   GPA: 4.27/4.3

    Graduated as the top student from the College of Natural Sciences (Valedictorian)

PUBLICATIONS

  • [NeurIPS 2025, Spotlight] Ψ-Sampler: Initial Particle Sampling for SMC-Based Inference-Time Reward Alignment in Score Models

    Taehoon Yoon*,   Yunhong Min*,   Kyeongmin Yeo*,  Minhyuk Sung (* equal contribution)

    | Website | Paper | arXiv | Code |

  • [NeurIPS 2025] Inference-Time Scaling for Flow Models via Stochastic Generation and Rollover Budget Forcing

    Jaihoon Kim*,  Taehoon Yoon*,   Jisung Hwang*,   Minhyuk Sung (* equal contribution)

    | Website | Paper | arXiv | Code |

  • [NeurIPS 2024] GrounDiT: Grounding Diffusion Transformers via Noisy Patch Transplantation

    Phillip Y. Lee*,  Taehoon Yoon*,   Minhyuk Sung (* equal contribution)

    | Website | Paper | arXiv | Code |

HONORS & AWARDS

  • Valedictorian, College of Natural Sciences, Sogang University   [Feb. 2024]
  • Dean's List, Sogang University   [Feb. 2023, Feb. 2020]
  • Top 1% Grade Award, Sogang University   [Sept. 2020, Mar. 2020, Mar. 2019]

RESEARCH EXPERIENCE

  • Graduate Researcher at Visual AI Lab, KAIST (Advisor, Prof. Minhyuk Sung)   [Mar.2024 - Present]
  • Research Intern at Visual AI Lab, KAIST (Advisor, Prof. Minhyuk Sung)   [Jan.2024 - Feb.2024]
  • Research Intern at CJ AI Center, CJ Group [hosted by Seoul National Univ. AI Institute] (Advisor, Dr. Seong Jong Ha)  [Jul.2023 - Aug. 2023]
  • Undergraduate Research Assistant at Visual Computing Lab, Sogang University (Advisor, Prof. Joo Ho Lee)   [Sept. 2022 - Apr. 2023]
  • Research Internship at BISPL Lab, KAIST AI graduate school (Advisor, Prof. Jong Chul Ye)   [Jul. 2022 - Sept. 2022]

WORK EXPERIENCE

  • Observatory & Education Assistant at Donga Science Observatory   [May.2018 - Feb. 2019]

Pinned Loading

  1. KAIST-Visual-AI-Group/Psi-Sampler KAIST-Visual-AI-Group/Psi-Sampler Public

    [NeurIPS 2025, Spotlight] Official code for Initial Particle Sampling for SMC-Based Inference-Time Reward Alignment in Score-Based Generative Models

    Python 13

  2. KAIST-Visual-AI-Group/Flow-Inference-Time-Scaling KAIST-Visual-AI-Group/Flow-Inference-Time-Scaling Public

    [NeurIPS 2025] Official code for Inference-Time Scaling for Flow Models via Stochastic Generation and Rollover Budget Forcing

    Python 69 1

  3. KAIST-Visual-AI-Group/GrounDiT KAIST-Visual-AI-Group/GrounDiT Public

    [NeurIPS 2024] Official Implementation of GrounDiT

    Jupyter Notebook 58 3

  4. Diffusion-Probabilistic-Models Diffusion-Probabilistic-Models Public

    PyTorch implementation for DDPM & DDIM

    Python 38 2

  5. SPH-Fluid-Simulation SPH-Fluid-Simulation Public

    Smoothed Particle Hydrodynamics implementation with Python

    Python 28 2

  6. CT-denoising-using-CycleGAN CT-denoising-using-CycleGAN Public

    AAPM Low Dose CT data denoising using Cycle GAN

    Jupyter Notebook 27 4