Hoyun Song (송호윤)
Hoyun Song (송호윤)
Embedded Files
I am a dedicated NLP/AI researcher focused on developing socially responsible AI, driven by the belief that high-quality, well-refined data is essential for robust and human-aligned model performance. I am currently working as an assistant professor in the Department of Artificial Intelligence at Dankook University. My core research interests include:
(1) Domain-Specific Modeling and Knowledge Integration for critical sectors such as mental healthcare and online safety, emphasizing the incorporation of domain expertise into modeling and data generation processes;
(2) Personalization and Human-Alignment to ensure language models mirror nuanced value preferences and maintain consistent personas; and
(3) Social Simulation and Behavior Analysis utilizing computational social science to assess how LLMs handle systemic roles, obligations, and social complexities.
These interests focus on creating evaluative benchmarks, knowledge-integrated modeling, and simulating complex institutional behaviors to steer the development of AI systems that aim to serve and benefit humanity.
Email: hoyun.song@dankook.ac.kr
Curriculum Vitae (CV): pdf
📖 Featured Publications (*: equal contribution)
Domain-Specific Modeling Knowledge Integration
MENTALBENCH: A Benchmark for Evaluating Psychiatric Diagnostic Capability of Large Language Models
Hoyun Song*, Migyeong Kang*, Jisu Shin, Jihyun Kim, Chanbi Park, Hangyeol Yoo, Jihyun An, Alice Oh, Jinyoung Han, KyungTae Lim
Under Review [paper]
Domain-Specific Modeling Knowledge Integration
MENTOR: A Reinforcement Learning Framework for Enabling Tool Use in Small Models via Teacher-Optimized Rewards
ChangSu Choi*, Hoyun Song*, Dongyeon Kim, WooHyeon Jung, Minkyung Cho, Sunjin Park, NohHyeob Bae, Seona Yu, KyungTae Lim
EMNLP 2026 [paper]
Domain-Specific Modeling Knowledge Integration
Does Rationale Quality Matter? Enhancing Mental Disorder Detection via Selective Reasoning Distillation
Hoyun Song, Huije Lee, Jisu Shin, Sukmin Cho, Changgeon Ko, and Jong C. Park
Findings of ACL 2025 [paper]
Human-aligned LLM
Spotting Out-of-Character Behavior: Atomic-Level Evaluation of Persona Fidelity in Open-Ended Generation
Jisu Shin, Juhyun Oh, Eunsu Kim, Hoyun Song, and Alice Oh
Findings of ACL 2025 [paper]
Human-aligned LLM
Different Bias Under Different Criteria: Assessing Bias in LLMs with a Fact-Based Approach
Changgeon Ko, Jisu Shin, Hoyun Song, Jeongyeon Seo, and Jong C. Park
SoLaR@NeurIPS 2024 [paper]
Social Simulation Behavior Analysis
RoleConflictBench: A Benchmark of Role Conflict Scenarios for Evaluating LLMs' Contextual Sensitivity
Jisu Shin, Hoyun Song, Juhyun Oh, Changgeon Ko, Eunsu Kim, Chani Jung, and Alice Oh
Findings of ACL 2026 [paper]
Social Simulation Behavior Analysis
Ask LLMs Directly, "What shapes your bias?": Measuring Social Bias in Large Language Models
Jisu Shin, Hoyun Song, Huije Lee, Soyeong Jeong, and Jong C. Park
Findings of ACL 2024 [paper]
💻 Work Experience
Assistant Professor @ Dankook University
2026.09 - Current
Department of Artificial Intelligence
Postdoctoral Researcher @ Korea Advanced Institute of Science and Technology (KAIST)
2025.08 - 2026.08
InnoCORE Postdoctoral Fellow
Research collaboration with Professor KyungTae Lim
🎓 Education
Korea Advanced Institute of Science and Technology (KAIST)
2019.03 - 2025.08
Advisor: Jong C. Park
Ph.D in School of Computing
Dissertation: Integration of Domain-Specific Knowledge for Mental Health Diagnostics
Korea Advanced Institute of Science and Technology (KAIST)
2017.03 - 2019.02
Advisor: Jong C. Park
M.S. in School of Computing
Thesis: Interpretable Depression Detection from Social Media using Hierarchical Attention Network with Depression Indicators
Dongguk University
2011.03 - 2017.02
Advisor: Eun Man Choi
BS in Computer Engineering (major), Business School (minor)
Page updated
Google Sites
Report abuse