Inchang Baek

I'm an integrated M.S.-Ph.D. candidate at the AI Graduate School of Gwangju Institute of Science and Technology (GIST), advised by Prof. Kyung-Joong Kim. My research sits at the intersection of reinforcement learning and human-AI interaction—specifically, developing methods that let RL agents receive natural-language instructions and produce outputs that align with human intent, using procedural content generation (PCG) as a structured testbed. For a detailed account of the research agenda, see my research statement.

Previously, I was a research intern at KRAFTON AI. I expect to graduate in February 2027.

Email / CV / Google Scholar / LinkedIn / Github

News

[Jul 2025] Invited to Dagstuhl Seminar 25292 — New Frontiers in AI for Game Design (Jul 13–18, 2025).
[Aug 2024] Visiting researcher at Game Innovation Lab, NYU with Julian Togelius (Aug – Nov 2024). Worked on automated reward generation for reinforcement learning using LLM reasoning capabilities.
[Feb 2023] Research intern at Game Innovation Lab, NYU (Director: Julian Togelius).

Publications

	Multiverse: Language-Conditioned Multi-Game Level Blending via Shared Representation I.-C. Baek, J. Jung, S.-H. Kim, G.-H. Hwang, K.-J. Kim Accepted at IEEE Conference on Games (CoG), 2026 (* Equal contribution) arXiv / code Multimodal
	Multi-Objective Instruction-Aware Representation Learning in Procedural Content Generation Reinforcement Learning S.-H. Kim, G.-H. Hwang, I.-C. Baek, S.-Y. Lee, K.-J. Kim Accepted at IEEE Conference on Games (CoG), 2026 arXiv RLMultimodal
	μCap: Instrumental Music Captions for Deaf and Hard-of-Hearing Individuals S. Ahn, I.-C. Baek, K.-J. Kim, K. N. Truong, J.-H. Hong ACM CHI Conference on Human Factors in Computing Systems (CHI), 2026 🏆 Best Paper Award (Top 1%) code LLM
	GPTalk: LLM-Based Virtual Companions for Metacognitive Growth in Self-Directed E-Learning Environments I.-T. Jung, C.-H. Lee, I.-C. Baek, D.-I. Oh, Y.-J. Choi, K.-J. Kim, D.-J. Kong, J.-H. Hong International Journal of Human-Computer Studies (IJHCS), 2026 paper LLM
	Human-Aligned Procedural Level Generation Reinforcement Learning via Text-Level-Sketch Shared Representation I.-C. Baek, S. Lee, S.-H. Kim, G. Hwang, K.-J. Kim Under revision at IEEE Transactions on Games, 2025 (* Equal contribution) arXiv RLMultimodal
	PCGRLLM: Large Language Model-Driven Reward Design for Procedural Content Generation Reinforcement Learning I.-C. Baek, S.-H. Kim, S. Earle, Z. Jiang, J.-H. Noh, J. Togelius, K.-J. Kim Under revision at IEEE Transactions on Games, 2025 arXiv RLLLM
	IPCGRL: Language-Instructed Reinforcement Learning for Procedural Level Generation I.-C. Baek, S.-H. Kim, S.-Y. Lee, D.-H. Lee, K.-J. Kim IEEE Conference on Games (CoG), 2025 (* Equal contribution) arXiv / paper RLMultimodal
	Seamless Tutorial: Contextual State Transition Generation Based on Player Internal Knowledge I.-C. Baek, T.-H. Park, K.-J. Kim IEEE Transactions on Games, 2025 paper / code / demo MCTS
	A Humanoid Visual-Tactile-Action Dataset for Contact-Rich Manipulation E.-J. Kwon, S.-W. Oh, I.-C. Baek, Y.-C. Park, G.-B. Kim, J.-Y. Moon, Y.-H. Choi, K.-J. Kim IROS 2025 Workshop on Robotic Fine Manipulation, 2025 arXiv IL
	ChatPCG: Large Language Model-Driven Reward Design for Procedural Content Generation I.-C. Baek, T.-H. Park, J.-H. Noh, C.-M. Bae, K.-J. Kim IEEE Conference on Games (CoG), 2024 arXiv / paper RLLLM
	RaidEnv: Exploring New Challenges in Automated Content Balancing for Boss Raid Games H.-C. Jeon, I.-C. Baek, C.-M. Bae, T. Park, W. You, T. Ha, H. Jung, J. Noh, S. Oh, K.-J. Kim IEEE Transactions on Games, 2023 (* Equal contribution) arXiv / paper / code RL
	Toward Cooperative Level Generation in Multiplayer Games: A User Study in Overcooked! I.-C. Baek, T.-G. Ha, T.-H. Park, K.-J. Kim IEEE Conference on Games (CoG), 2022 paper / code GA
	Turing Test Framework for Cooperative Games I.-C. Baek, T.-H. Park, T.-G. Ha, K.-J. Kim IEEE Conference on Games (CoG), 2022 paper RL
	A Swapping Target Q-value Technique for Data Augmentation in Offline Reinforcement Learning H.-T. Joo, I.-C. Baek, K.-J. Kim IEEE Access, vol. 10, 2022 paper RL
	Efficient Multi-Agent Reinforcement Learning Using Clustering for Many Agents I.-C. Baek, K.-J. Kim AIIDE-19 Workshop on Artificial Intelligence for Strategy Games, 2019 pdf RLMulti-Agent
	Web-Based Interface for Data Labeling in StarCraft I.-C. Baek, K.-J. Kim IEEE Conference on Computational Intelligence and Games (CIG), 2018 paper

Work Experience

KRAFTON AI — Research Intern
Aug 2025 – Oct 2025, Seoul, South Korea
Worked on LLM-based player agent policy generation for game environments.

Selected Honors

🏆 Best Paper Award (Top 1%), ACM CHI Conference on Human Factors in Computing Systems, 2026
🥈 2nd Place, The 2nd ChatGPT4PCG Competition, IEEE Conference on Games, 2024
🎓 GIST Graduate International Research Experience Fellowship (GIST-IREF), 2024

Selected Research Projects

Development of AI Players for Puzzle Game Automated Testing
PuzzleOne Studio (BitMango) (Industry-academia Project) | May 2022 – Jan 2023 | Team Leader
Gymnasium-based game simulator for commercial puzzle game; RL agents with categorical state representations

Development of AI-based Game Simulation Technology to Support Online Game Content Production
Korea Creative Content Agency (KOCCA) | Apr 2022 – Dec 2024
Unity-based multiplayer game gymnasium environment; LLM-driven reward generation for RL agents

Human-centered Game AI Basic Research Lab
National Research Foundation of Korea (NRF) | Jun 2021 – Feb 2024
Multiplayer game level generation using genetic algorithms; puzzle level generation & user study

Skills

Programming Languages: Python, C#, JavaScript
ML / RL Frameworks: PyTorch, JAX
Game Engines & Simulation: Unity
Infrastructure & Tools: Slurm, Docker

Website template from Jon Barron. Feel free to use the source code.