I'm a graduating undergraduate student at Georgia Tech majoring in Computer Science with concentrations in AI and Theory. I'm interested in ML and data visualization research and have publications in these areas.

Computer Science with concentrations in AI and Theory

Georgia Institute of Technology

*   Research under Professor Polo Chau
*   Co-led and co-first authored [_Transformer Explainer: Interactive Learning of Text-Generative Models_](https://poloclub.github.io/transformer-explainer/) (Hugging Face #1 paper of day; 125k+ users from 200+ countries; taught in Georgia Tech CSE6242 Data and Visual Analytics for 300 students per sem)
*   Presented at IEEE VIS ‘24 (top data visualization conference) as poster paper (Single Best Poster Award); currently under review at AAAI ‘25 Demo Program
*   Led backend development; extracted model data and converted PyTorch implementation to ONNX in order to run a pre-trained GPT-2 model live in browser
*   Used Figma to design an interactive visualization tool for explaining transformer architecture for non-experts

Research Assistant

*   Selected as one of 320 [scholarship recipients](https://oue.gatech.edu/node/3087) out of 2,358 international applicants to conduct research in Germany
*   Co-authored paper (currently under review at ICLR ‘25)
*   Executed SLURM scripts for training and experimenting on CLIP, to quantify and mitigate memorization in multimodal models
*   Contributed to the mathematical formulation of a novel definition for memorization in CLIP
*   Dissected and analyzed vision and language encoders of CLIP to study their respective roles in memorization, gaining insights into internal mechanisms of CLIP and comparing performance with uni-modal models, both supervised and self-supervised

*   First-authored paper on the environmental footprint of AI to develop a responsible AI guideline for promoting sustainability
*   Presented at AAAI/ACM Conference on AI, Ethics, and Society ‘24 (grant from President's Undergraduate Research Awards)
*   Research responsible data practices and ethics of human data annotation in collaboration with Google and DataWorks ATL
*   Used Pandas to load and process Google’s DICES dataset, retrieving and filtering data to create a data annotation guide
*   Presented and discussed papers on human-centered computing and AI in weekly meetings with 3 professors and 3 PhD students

*   TAed for CS 1331 (Object-Oriented Programming)
*   Led 2-hour recitation classes each week, instructing 50 students in fundamental object-oriented programming concepts
*   Conducted 3 hours of office hours each week, training students (out of 800 enrolled) in coding and debugging skills
*   Created and graded course materials including weekly programming assignments and exams in Java

Teaching Assistant

*   Increased dataset using the BRAT annotation tool, identifying and classifying self-disclosure phrases from 4,000+ Reddit posts
*   Improved cooking chatbot by chatting with crowdsourcing workers to get more natural dialogues over different recipes
*   Presented and discussed NLP papers in undergraduate reading groups

Selected as one of 320 [scholarship recipients](https://oue.gatech.edu/node/3087) out of 2,358 international applicants to conduct research in Germany

DAAD RISE Germany Scholarship

Received $1,000 grant to attend AIES 2024 conference to present my research

President's Undergraduate Research Travel Award

Full-ride merit scholarship given to the top 1% of applicants (40 students) to Georgia Tech

Grace Kim

#1 Video-Talent Marketplace

Ready to discover your perfect career?