Munyeong Kim | 김문영 (金汶暎)

Munyeong Kim | 김문영 (金汶暎)

he/him

Ph.D. student

Human-AI Agent Interaction and AI Memory

I am Munyeong Kim (김문영, 金汶暎), a Ph.D. student in the Department of Computer Science and Operations Research (DIRO) at Université de Montréal.

My research focuses on human-AI agent interaction, especially the memory systems that interactive agents build over time. I study what agents remember through interaction, how those memories can be refined, and how users can shape or design them.

Current Focus

Human-Computer Interaction, Human-AI interaction, shared memory between humans and AI agents, interfaces that support human productivity, agentic decision-making and human behavior simulation.

Find Me

Google Scholar, LinkedIn, X, Montréal HCI, Mila

CV

Click here. Current version: May 4, 2026.

Publications

  • Munyeong Kim, Michalis Famelis, and Ian Arawjo. "Metaphors for Memory: Charting a Design Space of AI Memory Tools and Interfaces." Proceedings of the 2026 ACM Designing Interactive Systems Conference (DIS '26). PDF DOI Slides Poster (JDIRO'26)
    About this paper

    This paper is for people thinking about how "memory" should actually work in interactive LLM-based agents and systems. We surveyed current AI memory tools and interfaces, then used metaphors to think through design possibilities that are not yet common in today's systems. We looked across a broad range of industry systems and academic prototypes, trying to capture what they have in common, where their current designs could be extended, and what questions are worth carrying forward. I hope it is useful for builders designing memory-related interactions, and for researchers looking for open questions in AI memory from a human-AI interaction perspective.

  • Priyan Vaithilingam, Munyeong Kim, Frida-Cecilia Acosta-Parenteau, Daniel Lee, Amine Mhedhbi, Elena L. Glassman, and Ian Arawjo. "Semantic Commit: Helping Users Update Intent Specifications for AI Memory at Scale." UIST 2025. DOI
  • Munyeong Kim and Sungsu Kim. "Generative AI in Mafia-like Game Simulation." arXiv preprint, 2023. Earlier archival/preprint version of the project. arXiv
  • Munyeong Kim. "GPTs in Mafia-like Game Simulation." CHI 2024 Student Research Competition. Conference version of the earlier project; research advised by Sungsu Kim. ACM Digital Library

For a fuller list, see Google Scholar.

Thesis

  • Munyeong Kim. "Designing Interaction and Management for AI Agent Memory: Design Space, Metaphors, and Interactive Conflict Management." M.Sc. thesis (mémoire), Université de Montréal, 2026. PDF

Talks

  • "Creating General User Models from Computer Use." Mila HCAI Reading Group, Mila - Quebec Artificial Intelligence Institute, Montréal, QC, Canada, August 5, 2025.
  • "Research Previews: Metaphors for Memory & dBlocks." Mila HCAI Reading Group, Mila - Quebec Artificial Intelligence Institute, Montréal, QC, Canada, February 18, 2026. Co-presented with Avinash Bhat (McGill) and Patrick Yung Kang Lee (U of T).

Volunteer

  • Student Volunteer (Photographer), ACM UIST 2025, Busan, Republic of Korea.