Woojung Song

M.S. in Data Science, Seoul National University (Sep 2025 ~)

B.S. in Information System, Hanyang University

Research Interests

I am interested in agents and value alignment. As AI agents become more integrated into everyday life, we need reliable ways to understand what values they hold and how they actually behave. On the value alignment side, I work on psychometric evaluation of LLMs: whether human questionnaires reliably measure model traits, how data contamination distorts those measurements, and how to build better evaluation instruments. On the agent side, I study how to make smaller models more capable tool users and how agents cope with realistic, uncooperative users.

I'm always happy to talk about shared interests. Feel free to reach out at opusdeisong@snu.ac.kr.

Publications

Full list on the Publications page.

Agents

Don't Adapt Small Language Models for Tools; Adapt Tool Schemas to the Models

ACL 2026 (Main)

Jonggeun Lee*, Woojung Song*, Jongwook Han*, Haesung Pyun, Yohan Jo  Co-first

Proposes PA-Tool, a training-free method that renames tool schema components to align with small language models' pretraining familiarity, achieving up to 17% improvement on tool-use benchmarks without model retraining.

Value Alignment

Human Psychometric Questionnaires Mischaracterize LLM Psychology: Evidence from Generation Behavior

Under Review, 2026

Woojung Song*, Dongmin Choi*, Yoonah Park, Jongwook Han, Eun-Ju Lee, Yohan Jo  Co-first

Shows that LLM psychological profiles from standard questionnaires diverge from profiles derived from actual generation behavior on real user queries.

Value Alignment

Quantifying Data Contamination in Psychometric Evaluations of LLMs

EACL 2026 (Findings)

Jongwook Han*, Woojung Song*, Jonggeun Lee*, Yohan Jo  Co-first

A systematic framework to measure data contamination across item memorization, evaluation memorization, and target score matching in psychometric LLM benchmarks.

Value Alignment

Value Portrait: Assessing Language Models' Values through Psychometrically and Ecologically Valid Items

ACL 2025 (Main)

Jongwook Han*, Dongmin Choi*, Woojung Song*, Eun-Ju Lee, Yohan Jo  Co-first

Proposes the Value Portrait framework for evaluating LLM values using items from real user-LLM interactions, finding across 44 LLMs that they emphasize Benevolence, Security, and Self-Direction.

Applied ML

Interpretable Prediction of Private Brand Purchases by Pet Type in E-Commerce for Consumer Behavior Analysis Using Real-World Transaction Data

PeerJ CS, 2025

Jaehyuk Lee*, Woojung Song*, Jina Kim, Eunchan Kim  Co-first