Kyuil Sim

M.S. Student, Graduate School of Industrial & Systems Engineering, KAIST

prof_pic_kyuil.jpg

I am an M.S. Student, Graduate School of Industrial & Systems Engineering, KAIST, advised by Prof. Jinkyoo Park in the System Intelligence Lab.

I received my B.S. in Industrial Engineering from Hanyang University. I worked with Prof. Jaewook Song at the Financial Innovation & AnalytiX - FINX Lab, focusing on graph-structured problems, particularly algorithm design and problem formulation. During an internship at SPACEMAP, I developed satellite collision probability algorithms and shortest-path algorithms for satellite communications.

I am broadly interested in leveraging Machine Learning for Combinatorial Optimization (ML4CO) to solve complex, real-world problems, with a particular focus on integrating ML with Mixed-Integer Linear Programming (MILP). Currently, my research centers on exploring and developing discrete diffusion models to tackle these optimization challenges.

news

May 2026 Two papers accepted to ICML 2026 Workshop SPIGM!
Sep 2025 Paper accepted to NeurIPS 2025 Workshop SPIGM!
Feb 2025 Started M.S. at the Graduate School of Industrial & Systems Engineering, KAIST, advised by Prof. Jinkyoo Park.

selected publications

  1. Solving Integer Linear Programming with Parallel Tempering
    Kyuil Sim, Sanghyeok Choi, and Jinkyoo Park
    arXiv preprint arxiv:2605.29366, 2026
  2. Aligning Few-Step Generative Models by Amortizing Sample-based Variational Inference
    Jaewoo Lee, Hyeongyu Kang, Dohyun Kim, and 9 more authors
    arXiv preprint arxiv:2605.26552, 2026
  3. Posterior Inference in Latent Space for Scalable Constrained Black-box Optimization
    Kiyoung Om*, Kyuil Sim*, Taeyoung Yun*, and 2 more authors
    arXiv preprint arXiv:2507.00480, 2025
    *Equal contribution.
  4. JSSE
    AstroLibrary: A library for real-time conjunction assessment and optimal collision avoidance
    Shawn SH Choi, Peter JH Ryu, Kyuil Sim, and 4 more authors
    Journal of Space Safety Engineering, 2024