# CSI-Agent Group

> CSI-Agent Group is a research group at Sungkyunkwan University (SKKU), Suwon, Korea, led by Prof. Honguk Woo. It works on embodied and physical AI, agentic AI, reinforcement learning, memory and neuro-symbolic reasoning, and real-world robot agents.

This file is generated from the website's data on every build. It indexes the public pages and summarises what they contain. For the complete member, publication, project and news lists in one document, read [llms-full.txt](https://csiagentgroup.com/llms-full.txt).

Last updated: 2026-10-09
Canonical site: https://csiagentgroup.com/

## Pages

- [Home](https://csiagentgroup.com/): Lab overview, research areas, recent news, recent demo videos, lab location and contact.
- [People](https://csiagentgroup.com/people/): Faculty, Ph.D. and M.S. students, undergraduate researchers, members on leave and alumni, with teams, research interests and Google Scholar links.
- [Projects](https://csiagentgroup.com/projects/): Current and completed research projects with periods, roles, topics and demo videos.
- [Publications](https://csiagentgroup.com/publications/): All papers by year with authors, venues, arXiv/OpenReview/DOI/code links and BibTeX.
- [Demos](https://csiagentgroup.com/demos/): Demo videos from the lab's research projects.
- [Photos](https://csiagentgroup.com/photos/): Photos from conferences, awards, workshops and lab life, by year.

## Quick facts

- Lab director: Prof. Honguk Woo (우홍욱), Professor, Sungkyunkwan University.
- Current members: 30 students and researchers (Ph.D. 15, M.S. 11, undergraduate 2, on leave 2); 18 alumni.
- Publications listed: 68 (2026: 15).
- Projects listed: 33 (16 ongoing).
- Location: Industry-Academia Cooperation Center 85561, 2066 Seobu-ro, Jangan-gu, Suwon, Gyeonggi-do, Korea. Weekdays 09:00–18:00 KST.

## Research areas

- Embodied and physical AI: agents that perceive, reason and act in physical and simulated environments.
- Agentic AI and reinforcement learning: adaptive decision-making for autonomous and collaborating agents.
- Memory and neuro-symbolic reasoning: knowledge, symbolic structure and reliable long-horizon reasoning.
- Robotics and real-world agents: robot manipulation, policy transfer and real-world robotic systems.

## Recent news

- 2026-09-25: Four Papers Accepted to NeurIPS 2026 (1 Oral, 1 Spotlight)
- 2026-08-20: Three Papers Accepted to EMNLP 2026 Main Conference and One to CIKM 2026
- 2026-06-18: Two Papers Accepted to IROS and ECCV 2026
- 2026-05-15: Three Papers Accepted to ICML 2026
- 2025-09-19: Three Papers Accepted to NeurIPS 2025 (2 Spotlight)

## Recent publications

- HeMeR: Heterogeneous Memory Reconciliation for Embodied Agents via Structured KV Reuse — Saehun Chun, Wonje Choi, Jinwoo Jang, TaeYoon Kwack, Honguk Woo. NeurIPS 2026 (oral)
- NeSyKC: Neurosymbolic Knowledge Compilation For Lifelong Learning Embodied Agents — Wonje Choi, Jooyoung Kim, Sera Choi, Younguk Song, Honguk Woo. NeurIPS 2026 (spotlight)
- Skill-Adaptive Noise Scheduling for Diffusion Policies — Woo Kyung Kim, Gwangpyo Yoo, Eunyoung Park, Honguk Woo. NeurIPS 2026
- Reconfiguring Procedural Knowledge for Compositional Robot Skill Adaptation — Daehee Lee, Dongsu Lee, Sanghyun Ahn, Minjong Yoo, Honguk Woo. NeurIPS 2026
- Visual Compliance via Executable Safety Rule Entailment — Jisoo Kim, TaeYoon Kwack, Jinwoo Jang, Honguk Woo. EMNLP 2026 — [arXiv](https://arxiv.org/abs/2609.18328)
- MA-Mesh: Resilient Multi-Agent Collaboration via Role-based Self-Healing — Junyong Lee, Woo Kyung Kim, Honguk Woo. EMNLP 2026
- ParentingLLM: Pacing Reasoning Maturity for Bidirectional Test-Time Scaling under Token Budgets — Seunghwan Jeong, Sugi Lee, Wei-Jin Park, Ikjun Yeom, Honguk Woo. EMNLP 2026
- Code-as-Auditor: Executable Compliance Reasoning via Regulation-to-Code — Jisoo Kim, TaeYoon Kwack, Jinwoo Jang, Woo Kyung Kim, Honguk Woo. CIKM 2026 — [arXiv](https://arxiv.org/abs/2609.19199)
- Multi-scale Mixture of World Models for Embodied Agents in Evolving Environments — Jinwoo Jang, Daniel J. Rho, Sihyung Yoon, Hyunsuk Cho, Honguk Woo. ECCV 2026 — [arXiv](https://arxiv.org/abs/2607.00457)
- RoboBRIDGE: A Modular Framework for Bridging Policies to Robust Real-World Robotic Agents — Sihyung Yoon, Minjong Yoo, Sanghyun Ahn, Seojeong Choi, Honguk Woo. IROS 2026 — [arXiv](https://arxiv.org/abs/2607.27881)
- Knothe-Rosenblatt Quantile Regression for Risk-sensitive Multi-objective Reinforcement Learning — Gwangpyo Yoo, Woo Kyung Kim, Honguk Woo. ICML 2026 — [OpenReview](https://openreview.net/forum?id=gGx6DyI6Xv)
- Functional Cache Grafting: Robust and Rapid Code-Policy Synthesis for Embodied Agents — Saehun Chun, Wonje Choi, Sera Choi, Sanghyun Ahn, Honguk Woo. ICML 2026 — [arXiv](https://arxiv.org/abs/2606.13097), [OpenReview](https://openreview.net/forum?id=hkmpUHQxgv)

## Ongoing projects

- Retrieval-Augmented Knowledge Search with LG Electronics (2026)
- Reconfigurable Lifelong Learning of VLA Models for Super-adaptive Embodied Agents (2026.11.20 (4 months))
- Resource-Intensive AI Technologies Based on Sustainable GPU Integrated Platforms (2026.04.01 – 2028.12.31)
- Development of Leading Talent in Medical Domain-Specific Generative AI (2026.04.01 – 2029.12.31)
- V2X Operation Optimization Using Deep Reinforcement Learning (2026.03.30 – 2027.03.29)
- Reactive Skill Diffusion for Real-world Embodied AGI (2026.03.01 – 2029.02.28)
- Autonomous Cluster Management (2025.12.30 – 2026.10.29)
- Test-Time Storage Scaling for LLM Serving (HiPER Phase 3) (2026.07.01 – 2029.06.30)
- Artificial Intelligence Graduate School Program(Sungkyunkwan University) (2019.04.01 – 2028.12.31)
- BK21 Education and Research Center for Sustainable Future Computing (2024.03.01 – 2027.08.31)
- AI Star Fellowship Support Program(Sungkyunkwan Univ.) (2025.07.01 - 2030.12.31)
- Accelerated Insight Reasoning via Continual Learning (2025.04.01 - 2029.12.31)
- Development of Open-Ended Alignment Fluxional AI for Ever-changing Environment and Value (2024.07.01 - 2031.12.31)
- Self-directed Multi-modal Intelligence for solving unknown, open domain problems (2022.04.01 - 2026.12.31)
- Adaptive Personality for Intelligent Agents (2022.04.01 - 2026.12.31)
- ICT Creative Consilience Program (2020.07.01 – 2029.12.31)

## Notes for AI agents

- The pages above are the source of truth; member lists, roles and papers change over time.
- Verify paper details against the Publications page and the linked arXiv, OpenReview, DOI or code pages.
- Do not infer private lab information, unpublished results, credentials or internal links from this file.

## Optional

- [llms-full.txt](https://csiagentgroup.com/llms-full.txt): every member, publication, project and news item in one Markdown document.
- [Sitemap](https://csiagentgroup.com/sitemap.xml)