Hyochan Chong (정효찬)
Research Engineer @ Samsung Research d7chong [at] gmail [dot] comAbout
Hi, I am a Research Engineer on the AI Systems team at Samsung Research.
My interests sit at the intersection of efficient LLMs and agentic systems. I am particularly interested in turning large models into systems that are faster, cheaper, and deployable under real hardware constraints.
Recently, I have been focusing on low-bit LLM quantization and inference, and multi-agent systems.
News
- Apr 2026 NanoQuant and RaBiT have been accepted to ICML 2026!
- Mar 2026 Gave a talk on LLM quantization at SqueezeBits
- Feb 2025 Graduating and returning to Samsung Research as a full-time research engineer
- Dec 2024 Won 1st place in the National AI Chip Competition, for optimizing NPU inference pipelines for Furiosa AI
- July 2024 Starting my internship at Samsung Research
Publications
NanoQuant: Efficient Sub-1-Bit Quantization of Large Language Models
ICML 2026
RaBiT: Residual-Aware Binarization Training for Accurate and Efficient LLMs
ICML 2026
(* indicates equal contribution)
Experience
| Samsung Research Research Engineer ㄴ LLM inference + quantization, agents |
Feb 2025 - Present |
Honors and Awards
1st place overall + 1st in Server Track, National AI Chip Competition, 2024 ($30,000 prize)
ㄴ Awarded by Furiosa AI and the Ministry of Science and ICT
United Nations Medal, United Nations, 2022
Peacekeeping Operations Medal, Korea Ministry of Defense, 2022
ㄴ Awarded by the Minister of Defense