Ayoung Lee (이아영)
Hi, I am a 3rd-year Ph.D. candidate in Computer Science and Engineering at University of Michigan. I am fortunate to be advised by Prof. Lu Wang.
My research focuses on evaluation and post-training of LLMs on nuanced, non-verifiable tasks:
I have designed a structure that makes decision-making in dilemma situations verifiable [2],
and introduced rubrics to assess and improve culture-aware moral decision-making [5] and coding agents [6].
I am also interested in multi-agent and human-AI collaboration:
I aim to build high-fidelity simulations and interaction strategies that enable LLM agents to
work effectively and safely in teams.
Feel free to contact me for collaboration or questions!
(Last updated on Aug 27th)
Recent News
Education
University of Michigan (08/2024 ~ Present)
Ph.D. in Computer Science and Engineering (Advisor: Prof. Lu Wang)
Seoul National University (03/2020 ~ 02/2024)
B.S. in Computer Science and Engineering
GPA: 4.18/4.3, Summa Cum Laude, Rank: 1st in the department
Seoul Science High School (03/2017 ~ 02/2020)
The school for gifted students in mathematics and science, teaching college-level math and science
GPA: 4.18/4.3
Publications
[6] Rubrics as Test Cases: Test-Grounded Execution-Free Verification of Coding Agents
Ayoung Lee, Justus Mattern
In preparation (submitting to ARR October 2026)
[5] MET: Theory-Grounded and Culture-Aware Multilingual Moral Reasoning
Ayoung Lee, Ryan Sungmo Kwon, Yunxiang Zhang, Yuxuan Liu, Peter Railton, Lu Wang
COLM 2026 [paper]
[4] LiveOIBench: Can Large Language Models Outperform Human Contestants in Informatics Olympiads?
Kaijian Zou, Aaron Xiong, Yunxiang Zhang, Xinliang Frederick Zhang, Yueqi Ren, Jirong Yang, Ayoung Lee, Shitanshu Bhushan, Lu Wang
ICML 2026 [paper]
[3] Logit Arithmetic Elicits Long Reasoning Capabilities Without Training
Yunxiang Zhang, Muhammad Khalifa, Lechen Zhang, Xin Liu, Ayoung Lee, Xinliang Frederick Zhang, Farima Fatahi Bayat, Lu Wang
Findings of ACL 2026 [paper]
[2] CLASH: Evaluating Language Models on Judging High-Stakes Dilemmas from Multiple Perspectives
Ayoung Lee, Ryan Sungmo Kwon, Peter Railton, Lu Wang
ICLR 2026 [paper]
Honors and Awards (Selected)
Best Presentation Award, NLP @ Michigan Day (03/2025)
Kwanjeong Educational Foundation Scholarship (08/2024 ~ Present)
Excellent Bachelor's Thesis Presentation Award (02/2024)
Outstanding Tutor Award (08/2021)
Presidential Science Scholarship (03/2020 ~ 02/2024)
Work Experience
Proximal (05/2026 ~ 08/2026)
Research Intern (Host: Justus Mattern)Naver Cloud HealthCare AI (03/2024 ~ 06/2024)
Research Intern (Host: Heeyoung Kwak)Language and Data Intelligence Lab (03/2023 ~ 02/2024)
Undergraduate Research Intern (Advisor: Prof. Seung-won Hwang)DeepMetrics (06/2022 ~ 01/2023)
Machine Learning Engineering Intern (Advisor: Prof. Hyun Oh Song)Samsung Device Solutions Memory (07/2021 ~ 08/2021)
Software InternTalks
Injecting Math Reasoning Abilities in Language Models @ Deepest, Seoul National University (06/2024)
Study Abroad Presentation for Students @ Seoul Science High School (05/2024)
Teaching
(EECS 497) Human-Centered Software Design and Development @ University of Michigan (01/2026 ~ 04/2026)
Graduate Student Instructor
(M1522.000700) Logic Design @ Seoul National University (09/2021 ~ 12/2021)
Teaching Assistant
(034.005) Foundation of Physics 1 @ Seoul National University (03/2021 ~ 08/2021)
Peer Tutor
(M1522.000700) Logic Design @ Seoul National University (03/2021 ~ 08/2021)
Tutor