Jaesik Yoon

ML Developer at SAP and Ph.D student at KAIST

jaesik_pic.png

I’m a senior machine learning developer at SAP. I work on the development of intelligent agents on SAP’s products, especially on SAP HANA Cloud.

Besides, I’m a Ph.D. student at SoC at KAIST under the supervision of Prof. Sungjin Ahn. My research interests are in the diverse inductive biases to extend the generality of AI.

In the past, I worked in Samsung Electronics and graduated from KAIST.

news

Aug 21, 2025 Our code for MCTD and Fast-MCTD is released, and you can find it here!
Jun 12, 2025 Our paper Fast Monte Carlo Tree Diffusion: 100x Speedup via Parallel Sparse Planning is released on Arxiv!
May 21, 2025 Our paper Adaptive Cyclic Diffusion for Inference Scaling is released on Arxiv!

latest posts

Jun 16, 2025 Inference-Time Scaling with Verifiers: Democratizing AI Reasoning

selected publications

  1. Monte Carlo Tree Diffusion for System 2 Planning
    Jaesik Yoon, Hyeonseo Cho, Doojin Baek, and 2 more authors
    In International Conference on Machine Learning, 2025
    Spotlight (top 2.6% of submissions)
    Page PDF Code
  2. Spatially-Aware Transformers for Embodied Agents
    Junmo Cho*, Jaesik Yoon*, and Sungjin Ahn
    In International Conference on Learning Representations, 2024
    *Equal contribution, Spotlight (top 5% of submissions)
    Page PDF
  3. TransDreamer: Reinforcement Learning with Transformer World Models
    Chang Chen*, Yi-Fu Wu*, Jaesik Yoon*, and 1 more author
    In Deep RL Workshop NeurIPS, 2021
    * Equal contribution
    PDF
  4. Sequential neural processes
    Gautam Singh*, Jaesik Yoon*, Youngsung Son, and 1 more author
    Advances in Neural Information Processing Systems, 2019
    * Equal contribution, Spotlight (top 2% of submissions)
    Page PDF Code
  5. Bayesian model-agnostic meta-learning
    Jaesik Yoon*, Taesup Kim*, Ousmane Dia, and 3 more authors
    Advances in neural information processing systems, 2018
    * Equal contribution, Spotlight (top 3% of submissions)
    PDF Code