Skip to content
View delaykimm's full-sized avatar

Block or report delaykimm

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Maximum 250 characters. Please don't include any personal information such as legal names or email addresses. Markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
delaykimm/README.md

Hi there, I'm Jiyun Kim (Grace)👋

🎓Education

  • Computer Science & Engineering @ Korea University (Mar 2022 - Present)
  • Brain and Cognitive Sciences @ Korea University (Mar 2024 - Present)
  • Exchange Student (CS) @ The University of Texas at Austin (Jan 2025 - May 2025)

🔭 Research Experience

Undergraduate Intern, Data Science & Artificial Intelligence Laboratory (Advisor: Prof. Chanyoung Park), Korea Advanced Institute of Science and Technology (KAIST) Dec 2025 – Feb 2026

  • Reviewed prior publications from the lab and conducted independent research on diffusion-based VLA models.

Undergraduate Researcher, Data & Adaptive Intelligence Systems Lab (Advisor: Prof. Susik Yoon), Korea University
July 2024 – Dec 2025

  • Conducted independent research on time-series representation learning with large language models (LLMs).

🔭 Research Projects

Numerical-Token-Grounded Time-Series & Textual Embedding Alignment for Forecasting
June 2025 – Dec 2025
(Individually led research conducted at Data & Adaptive Intelligence Systems Lab under the supervision of Prof. Susik Yoon, Korea University)

  • Identified that existing time-series–text methods rely on naive channel-wise attention without explicit alignment criteria.
  • Proposed a numerical token–based alignment framework by converting time-series values into natural-language descriptions.
  • Aligned timestamp-level time-series embeddings with number-aware textual token embeddings, enabling fine-grained and interpretable cross-modal representations.
  • Validated the framework through basic experiments on ETTh1 and ILI datasets.
  • Manuscript in preparation.

🔭 Research Interests

  • Visual Perception & Scene Understanding
  • Interpretable & Structured Representation Learning
  • Multimodal Representation Learning
  • Visual Language Action Models (VLAs)
  • Autonomous Driving Perception

🛠️ Tech Stack

Python PyTorch NumPy Pandas scikit-learn Transformers HuggingFace HDF5 Matplotlib Linux GitHub

📫 Contact

“Keep it simple, but meaningful”

Pinned Loading

  1. time-text-alignment time-text-alignment Public

    Python

  2. AIKU-Official/aiku-24-1-should-i-buy AIKU-Official/aiku-24-1-should-i-buy Public

    Python

  3. AIKU-Official/aiku-24-2-video_retrieval AIKU-Official/aiku-24-2-video_retrieval Public

    Python 1

  4. AIKU-Official/baseball-highlight-generator AIKU-Official/baseball-highlight-generator Public

    AIKU 2024-Summer Project: Baseball Highlight Generator

    Jupyter Notebook

  5. 20252R0136DATA30400 20252R0136DATA30400 Public

    final project : hierarchical multi-label text classification

    Python

  6. sungjaep11/nineup sungjaep11/nineup Public

    TypeScript