About

I am a Ph.D. student in the Interdisciplinary Program in Artificial Intelligence at Seoul National University, advised by Prof. Kyomin Jung. My work focuses on creative, diverse, and long document generation, multimodal generation, self-improving AI agents, and AI for scientific agents.

I received my B.S. in Electrical and Computer Engineering from Seoul National University in 2024. I am also a research intern at LG AI Research, where I work on FigureCode, a benchmark for generating and evaluating scientific architectural figures as editable HTML/SVG source code.

Recent News

  • 2026 Submitted FigureCode: Generating and Evaluating Scientific Architectural Figures as Editable HTML/SVG Source Code to May ARR.
  • 2026 A Universal Avoidance Method for Diverse Multi-branch Generation accepted to ACL 2026.
  • 2026 Fine-grained Story Visualization Evaluation via Diversified Storylines accepted to ACL 2026.
  • 2025 Avoidance Decoding for Diverse Multi-Branch Story Generation accepted to EMNLP 2025.
  • 2025 A Character-Centric Creative Story Generation via Imagination accepted to ACL 2025.
  • 2025 MultiActor-Audiobook: Zero-Shot Audiobook Generation with Faces and Voices of Multiple Speakers accepted to Interspeech 2025 as an oral presentation.
  • 2024 LongStory: Coherent, Complete and Length Controlled Long Story Generation accepted to PAKDD 2024 as an oral presentation.

Research Interests

  • Creative, diverse, and long document generation: controllable long-form generation, imagination-based generation, and multi-branch decoding.
  • Multimodal generation: systems that combine text, faces, and voices for richer content creation.
  • Self-improving AI agents: agents that can autonomously revise their own system architecture and behavior.
  • AI for scientific agents: AI systems that assist scientific research workflows, reason over scientific artifacts, and support the generation, evaluation, and refinement of research outputs beyond a single benchmark or artifact type.

Publications

* indicates primary contribution. Links point to arXiv pages when available.

  1. LongStory: Coherent, Complete and Length Controlled Long Story Generation
    Kyeongman Park*, Nakyeong Yang, and Kyomin Jung.
    PAKDD 2024, Oral Presentation. arXiv · PDF
  2. MultiActor-Audiobook: Zero-Shot Audiobook Generation with Faces and Voices of Multiple Speakers
    Kyeongman Park*, Seongho Joo, and Kyomin Jung.
    Interspeech 2025, Oral Presentation. arXiv · PDF
  3. A Character-Centric Creative Story Generation via Imagination
    Kyeongman Park*, Minbeom Kim, and Kyomin Jung.
    ACL 2025, Poster Presentation. arXiv · PDF
  4. Avoidance Decoding for Diverse Multi-Branch Story Generation
    Kyeongman Park*, Nakyeong Yang, and Kyomin Jung.
    EMNLP 2025, Poster Presentation. arXiv · PDF
  5. A Universal Avoidance Method for Diverse Multi-branch Generation
    Kyeongman Park*, Minha Jhang, and Kyomin Jung.
    ACL 2026, Poster Presentation. arXiv · PDF
  6. Fine-grained Story Visualization Evaluation via Diversified Storylines
    Minha Jhang*, Kyeongman Park, and Kyomin Jung.
    ACL 2026, Poster Presentation. arXiv link coming soon

Academic Service

  • Reviewer, ACL Rolling Review (ARR), Oct. 2025 and Feb. 2026.

CV

Download CV Updated June 2026.

Contact

Email: zzangmane@snu.ac.kr
LinkedIn: here
Lab: Machine Intelligence Lab, Seoul National University