I am interested in optimizing AI systems, particularly in AI compression and hardware acceleration.

Education

Research Focus

Mixed Precision Quantization

Optimizing bit-width per layer (Mixed Granularity)

Any Precision Quantization

Dynamic adaptation from High to Low Precision

Industrial Experiences

• Undergraduate Researcher, Sungkyunkwan University (Dec 2024 - Present)
  • Architectures for computationally efficient neural systems employing precision-scalable representations
  • Methodologies for robust optimization of models under extreme precision constraints
• AI Software Engineer, (주)선도소프트 (Freelance) (Jul 2024 - Sep 2024)
  • Develop LLM platform
  • Build RAG (Retrieval-Augmented Generation) system for large data
  • Implement quantization techniques for model efficiency
  • Optimize LLM serving for high-performance deployment
• Software Developer, Metabora (Internship) (Jan 2024 - Feb 2024)
  • Java, ERD

Publications

MSQ: Memory-Efficient Bit Sparsification Quantization

Seokho Han*, Seoyeon Yoon*, Jinhee Kim, Dongwei Wang, Kang Eun Jeon†, Huanrui Yang†, Jong Hwan Ko†

Published in ICCV 2025, 2025

* equal contribution, † corresponding author

Recommended citation: https://www.arxiv.org/pdf/2507.22349