Saehun Chun

Saehun Chun

Efficient AI Systems · Foundation Models · Embodied AI

I am an integrated M.S./Ph.D. student at Sungkyunkwan University in CSI Agent Group. My research focuses on efficient AI systems: inference, memory, and computation reuse for foundation models and interactive multimodal agents.

Current research

Efficient AI Systems
Memory, caching, and computation reuse.
Robot Foundation Models
Efficient multimodal inference and serving.
Embodied AI & Robot Learning
VLA models and interactive embodied agents.

Publications

All publications →

  1. HeMeR: Heterogeneous Memory Reconciliation for Embodied Agents via Structured KV Reuse

    Saehun Chun, Wonje Choi, Jinwoo Jang, TaeYoon Kwack, and Honguk Woo*

    NeurIPS 2026 · Oral

  2. Functional Cache Grafting: Robust and Rapid Code-Policy Synthesis for Embodied Agents

    Saehun Chun, Wonje Choi, Sera Choi, Sanghyun Ahn, and Honguk Woo*

    ICML 2026 · Poster / Paper

  3. Efficient Skill Grounding via Code Refactoring with Small Language Models

    Sera Choi, Wonje Choi, Saehun Chun, Daehee Lee, Chaeun Lee, Jooyoung Kim, and Honguk Woo*

    ICML 2026 · Poster / Paper