Graduate Researcher

Seohyun Lee

M.S. Researcher, Elmore Family School of Electrical & Computer Engineering, Purdue University

Hello! I am an M.S. researcher at the ION Lab in Purdue's Elmore Family School of ECE (Electrical and Computer Engineering), advised by Prof. Christopher G. Brinton. I am currently actively seeking Ph.D. opportunities in self-play learning and/or trustworthy and fair learning.

Portrait of Seohyun Lee
01

Research

My work currently centers on reliable, fair, and trustworthy machine learning and self-learning models, with the most experience within a Federated Learning (FL) setting, where models are trained across many distributed clients without sharing their raw data. I'm interested in how ML systems can stay robust and equitable, in conjunction with self-play learning as a way for models to keep improving from their own experience.

Federated Learning Foundation & World Models Edge / Distributed ML Self-Play & RL Personalization / Fairness Adversarial Robustness
02

Publications & Preprints

[1]

Self-Play Enhancement via Advantage-Weighted Refinement in Online Federated LLM Fine-Tuning with Real-Time FeedbackPreprint

S. Lee, W. Fang, D. Han, S. Hosseinalipour, C. G. Brinton

arXiv preprint arXiv:2605.07977, 2026

[2]

TAP: Two-Stage Adaptive Personalization of Multi-Task and Multi-Modal Foundation Models in Federated LearningPreprint

S. Lee, W. Fang, D. Han, S. Hosseinalipour, C. G. Brinton

arXiv preprint arXiv:2509.26524, 2026

[3]

Cooperative Decentralized Backdoor Attacks on Vertical Federated LearningIEEE/ACM ToN

S. Lee, W. Fang, A. B. Das, S. Hosseinalipour, D. J. Love, C. G. Brinton

IEEE/ACM Transactions on Networking, 2025

[4]

Smart Information Exchange for Unsupervised Federated Learning via Reinforcement LearningIEEE ICC 2024

S. Lee, A. B. Das, S. Wagle, C. G. Brinton

ICC 2024 — IEEE International Conference on Communications, pp. 3494–3499

[5]

Detecting and Mitigating Backdoor Attacks in OTA-FL Systems: A Two-Stage Robust Aggregation SchemePreprint

X. Ma, S. Lee, T. Kim, C. G. Brinton

arXiv preprint arXiv:2605.19253, 2026

03

Research Experience

May 2025 — Present
West Lafayette, IN

Graduate Research Assistant

ION Lab · Elmore Family School of Electrical & Computer Engineering, Purdue University

As a fully-funded M.S. researcher, my focus lies in developing algorithms for two main thrusts: (i) self-play learning and (ii) ensuring foundation models are fair, trustworthy, and reliable. In particular, I have developed personalized foundation models that work across multiple modalities and tasks, and on online self-play learning--directions where our methods have been beating the state of the art by roughly 4–8%.

May 2023 — May 2025
West Lafayette, IN

Research Member

ION Lab · Elmore Family School of Electrical & Computer Engineering, Purdue University

I joined the ION Lab during my undergraduate, working alongside PhD and postdoctoral researchers on edge learning. My focus was on two of the harder problems in Federated Learning: handling data heterogeneity across clients and adversarial backdoor attacks, where the approaches I develop outperformed the prior state of the art by 5–30%.

04

Teaching

Aug 2023 — May 2025
300 students · TDM 201/202

Teaching Assistant — The Data Mine Seminar II

Purdue University

I helped run this data-science seminar for a class of around 300 students, covering Python, R, bash, and SQL, with topics ranging from plotting and NumPy/Pandas to filtering data and exploring it from the command line. Week to week, I held office hours, wrote up assignment solutions, and graded coursework, and I stepped in to support the 900-student intro version (TDM 101) when the staff needed extra hands.

Aug 2022 — Dec 2022
400 students · CS 240

Teaching Assistant — Programming in C

Purdue University

This was an introductory C course for about 400 students, covering the fundamentals: file I/O, memory layout and management, and pointers. In addition, guidance for practical tools like SSH and Vim. I ran two weekly lab sessions of 30–40 students each, wrote and graded the post-lab quizzes, and proctored exams.

05

Education

M.S., Electrical & Computer Engineering

Purdue University, West Lafayette, IN

Aug 2025 — May 2027GPA 3.90 / 4.00

B.S. Computer Science · B.A. Minor, Japanese

Purdue University · John Martinson Honors College

Aug 2021 — May 2025GPA 3.90 / 4.00
06

Technical Skills

Languages

PythonJavaC / C++ RJavaScriptSQL MongoDBNeo4jARM

ML Frameworks & Libraries

PyTorchTensorFlowKeras scikit-learnHuggingFaceOpenAI Gymnasium NumPyPandasOpenCV Matplotlibseaborn

Web & Data

ReactNode.jsMySQL PostgreSQL

Tools & Languages Spoken

GitVimVS Code Google Colab EnglishKoreanJapanese
07

Service & Talks

Reviewer Service

  • IEEE/ACM Transactions on Networking (ToN)
  • Neural Information Processing Systems (NeurIPS)

Invited Presentations

  • “Smart Information Exchange for Unsupervised Federated Learning via Reinforcement Learning,” IEEE ICC, Denver, USA — June 11, 2024.