# CSI-Agent Group

> CSI-Agent Group is a research group at Sungkyunkwan University focused on embodied intelligence, agentic AI, reinforcement learning, memory, neuro-symbolic reasoning, and real-world robotic agents.

This file helps LLMs and browser agents find the main public pages on the CSI-Agent Group website. It is a curated index, not a sitemap. Only public website content is included.

Last updated: 2026-08-10
Canonical site: https://csiagentgroup.com/

## Site Overview

- [Home](https://csiagentgroup.com/): Lab overview, research areas, recent news, recent demos, location, and contact information.
- [People](https://csiagentgroup.com/people/): Faculty, Ph.D. students, master's students, undergraduate members, alumni, research teams, and public academic profile links.
- [Projects](https://csiagentgroup.com/projects/): Current and completed research projects, including project periods, participation roles, topics, tags, descriptions, and demo videos when available.
- [Publications](https://csiagentgroup.com/publications/): Research publications by year with titles, authors, venues, links to papers, OpenReview pages, arXiv pages, code, DOI entries, and BibTeX where available.

## Research Areas

- [Embodied and Physical AI](https://csiagentgroup.com/): Agents that perceive, reason, and act in physical and simulated environments.
- [Memory and Neuro-Symbolic Reasoning](https://csiagentgroup.com/): Memory, knowledge, symbolic structure, and reliable long-horizon reasoning.
- [Reinforcement Learning and Agentic AI](https://csiagentgroup.com/): Adaptive decision-making methods for autonomous and collaborative AI agents.
- [Robotics and Real-World Agents](https://csiagentgroup.com/projects/): Projects involving embodied agents, robot manipulation, robust policy transfer, and real-world robotic systems.

## Key Public Pages for Common Tasks

- [Find lab members](https://csiagentgroup.com/people/): Use this page for current people, roles, teams, advisors, public profiles, and research interests.
- [Find active research projects](https://csiagentgroup.com/projects/): Use this page for project summaries, status, periods, and topic tags.
- [Find papers and citations](https://csiagentgroup.com/publications/): Use this page for publication metadata, external paper links, code links, and BibTeX.
- [Find recent announcements](https://csiagentgroup.com/): Use the News section on the home page for recent accepted papers and lab updates.
- [Find lab location and contact](https://csiagentgroup.com/): Use the Visit CSI-Agent Group section on the home page for address, hours, and contact email.

## Recent Publication Themes

- Embodied agents in evolving or dynamic environments.
- Robust real-world robotic agents and policy bridging.
- Risk-sensitive and multi-objective reinforcement learning.
- Code-policy synthesis, code refactoring, and skill grounding for embodied agents.
- Neuro-symbolic planning, proceduralization, and continual task adaptation.
- World models, test-time adaptation, and skill incremental learning.

## Recent Project Themes

- AI Star Fellowship support for Human-AI companion collaboration.
- Continual learning for accelerated insight reasoning.
- Open-ended alignment for non-stationary environments and values.
- Self-directed multimodal intelligence for unknown open-domain problems.
- Adaptive personality for intelligent agents.
- Federated learning, edge-server inference, drone AI, storage systems, and AI application frameworks.

## Notes for AI Agents

- Prefer the public pages listed above as the source of truth.
- Do not infer private lab information, unpublished results, credentials, or internal links from this file.
- When answering questions about papers, verify details on the Publications page and follow external links such as arXiv, DOI, OpenReview, or code repositories when present.
- When answering questions about people, use the People page because member lists and roles may change.