Yejin Kim

Self-motivated and passionate graduate student researcher at ITML at Yonsei University, South Korea. Regardless of type of task, I try to figure out the essential principle from the problem and make a scalable solution.

yjkim.stat@yonsei.ac.kr

Interested in

Representaion Learning based on Information Theroy
Theory and applications of foundation models in language model(e.g. BERT, GPT)
Conversational AI and LLM based Evaluation

Past Research Topics

[Research] ERD : A Framework for Improving LLM Reasoning for Cognitive Distortion Classification
NAACL '24 Clinical Workshop
[arxiv Link]

[Intern] Text Embedding Model with Mistral 7B :Research Intern @ Linq
January 2024 ~ February 2024
[Post Link]

[Research] Edge-aware aggregation for deep Graph Neural Network
March 2023 ~ June 2023
[Github Link]

[Intern] Implement Neurocard which is Data-driven query optimization model on several environments and make API @ SAP Labs Korea
July 2022 ~ June 2023

[Paper] Data Importance-Aware Scheduling for Exploiting Diversity of Distributed Data
July 2021 ~ June 2022
[ Paper Link]

Project Topic

[Algorithm] Data-driven evaluation system on reading ability of elementary schoold students using eye tracking data @ Visual Camp corp
Dec 2022
(Public) [Github LINK]

[NLP] Framework for Text Style Transfer: from data processing to implmentation of Style LM model
Dec 2022
[Github LINK]

[Recommendation System] Implementation of Deep-learning based recommendation service on My-Data infrastructure
Dec 2022
[paper LINK]

[CV] implmentation of mobile application service of Deep-learning based Object Detection model
Dec 2022
[Github LINK]

[CV] Recommendation based on Instance Segmentation and Data Embedding
May 2022
[Github LINK]

[Clustering] Data-driven Social problem solution Competition
Sep 2021
[Github LINK]