Minseo Kim

Undergraduate Student, CSE, Seoul National University

Hi, I’m an undergraduate student at Seoul National University majoring in Computer Science and Engineering. I am broadly interested in efficient LLM inference and post-training. During my undergraduate years, I was a visiting researcher at Berkeley AI Research (BAIR), advised by Prof. Kurt Keutzer and Dr. Amir Gholami. Previously, I worked at the Architecture and Code Optimization Lab (ARC Lab) at Seoul National University, advised by Prof. Jae W. Lee.

Incoming UC Berkeley CS PhD student (Fall 2026).

prof_pic.png

news

Jan 27, 2026 Our paper on accelerating DLM inference is accepted to MLSys 2026. (TogetherAI blog)
Jan 19, 2026 I’m joining FuriosaAI as an AI Algorithm Research Intern!
Jan 08, 2026 I gave a 1-hour online seminar at Cerebras with Coleman, presenting our work on DLMs.

selected publications

  1. ICML AdaptFM
    efficientrollout.png
    EfficientRollout: System-Aware Self-Speculative Decoding for RL Rollouts
    Minseo Kim*, Minjae Lee*, Seunghyuk Oh, and 7 more authors
    ICML 2026 AdaptFM Workshop, Extended version under review , 2026
    arXiv
  2. ISCA MLArchSys
    llminaflash.png
    LLM Inference in a Flash!
    Sebastian Zhao*, Minseo Kim*, Coleman Hooper*, and 5 more authors
    ISCA 2026 MLArchSys Workshop (Oral), 2026
  3. MLSys
    cdlm.png
    CDLM: Consistency Diffusion Language Models for Faster Sampling
    Minseo Kim, Chenfeng Xu, Coleman Hooper, and 5 more authors
    In Proceedings of the 9th MLSys Conference, 2026
    arXiv Code
  4. beyond.png
    Beyond Next-Token Prediction: A Performance Characterization of Diffusion versus Autoregressive Language Models
    Minseo Kim, Coleman Hooper, Aditya Tomar, and 5 more authors
    arXiv Preprint , 2025
    arXiv