Minchan Jeong

AI Fellow, Center for AI and Natural Sciences, KIAS

Photo of Minchan Jeong

I'm an AI Fellow at the Center for AI and Natural Sciences, KIAS. I finished my PhD at KAIST in 2026, advised by Se-Young Yun (thesis: Operator Spectral Estimation Beyond IID Data). I build neural methods that recover an operator's leading spectrum with provable guarantees, for stochastic dynamical systems and for the excited states of quantum many-body systems. Two results anchor this, both with Jon Ryu: a parametric SVD of the Koopman operator (NeurIPS 2025) and low-rank variational Monte Carlo for quantum excited states (NestedLoRA-VMC).

Earlier I built weather ML systems at scale: a terabyte-scale data pipeline, then the JAX inference engine behind an operational GraphCast deployment at KMA.

Education

Selected publications

  1. Variational Monte Carlo for Quantum Excited States via Nested Low-Rank Approximation
    M. Jeong*, J. J. Ryu*, S.-Y. Yun, G. W. Wornell
    2026 . Manuscript under review, 2026.
    quantumspectral
  2. From Belief Entrenchment to Robust Reasoning in LLM Agents
    J. Oh*, M. Jeong*, J. Ko, S.-Y. Yun
    2026 . To appear in Transactions of the Association for Computational Linguistics (TACL).
    llm
  3. Efficient Parametric SVD of Koopman Operator for Stochastic Dynamical Systems
    M. Jeong*, J. J. Ryu*, S.-Y. Yun, G. W. Wornell
    Advances in Neural Information Processing Systems, 2025, pp. 25564–25600
    spectral
  4. Robust Streaming PCA
    D. Bienstock*, M. Jeong*, A. Shukla*, S.-Y. Yun*
    Advances in Neural Information Processing Systems, 2022, pp. 4231–4243
    spectral
  5. Preservation of the Global Knowledge by Not-True Distillation in Federated Learning
    G. Lee*, M. Jeong*, Y. Shin, S. Bae, S.-Y. Yun
    Advances in Neural Information Processing Systems, 2022, pp. 38461–38474
    federated learning

Full publication list →

Recent

Contact