Image

Dasol Choi

Principal Researcher, AIM Intelligence

M.S., Yonsei University

dasolchoi@yonsei.ac.kr

About

Hello! I'm a Principal Researcher at AIM Intelligence, where I lead evaluation work on the safety and reliability of language and multimodal models. I'm also an active researcher at HAE-RAE, an open-source community focused on Korean NLP, and I completed my M.S. at Yonsei University.

My research spans safety evaluation, alignment, bias mitigation, and adversarial robustness across multimodal and language models. The through-line is a single question: how much can we actually trust the AI systems we increasingly rely on, and how do we make them more dependable in real-world use?

News

Education

Yonsei UniversityMar. 2024 – Feb. 2026

M.S. in Digital Analytics, College of Computing

Kyung Hee UniversityMar. 2013 – Feb. 2019

B.A. in Japanese Language (Major) & B.A. in Physical Education (Double Major)

Publications

(* indicates equal contribution, † corresponding author)

For the complete list, see my Google Scholar.

When Cars Have Stereotypes: Auditing Demographic Bias in Objects from Text-to-Image Models

Dasol Choi, Jihwan Lee, Minjae Lee, Minsuk Kahng

ECCV 2026

Paper

COMPASS: A Framework for Evaluating Organization-Specific Policy Alignment in LLMs

Dasol Choi*, DongGeon Lee*, Brigitta Jesica Kartono*, Helena Berndt, Haon Park, Hwanjo Yu, Minsuk Kahng

ACL 2026

Paper

What Users Leave Unsaid: Under-Specified Queries Limit Vision-Language Models

Dasol Choi*, Guijin Son*, Hanwool Lee*, Minhyuk Kim, Hyunwoo Ko, Teabin Lim, Eungyeol Ahn, Jungwhan Kim, Seunghyeok Hong, Youngsook Song

ACL 2026 Findings

Paper

Distribution-Level Feature Distancing for Machine Unlearning: Towards a Better Trade-off Between Model Utility and Forgetting

Dasol Choi, Dongbin Na

AAAI 2025

Paper

Better Safe Than Sorry? Overreaction Problem of Vision Language Models in Visual Emergency Recognition

Dasol Choi, Seunghyun Lee, Youngsook Song

WACV 2026

Paper

LLM-Based Medical Document Evaluation: Integrating Human Expert Insights

Junhyuk Seo*, Dasol Choi*, Wonchul Cha, Taerim Kim

MedInfo 2025 (Best Student Paper Award)

Paper

Redefining Evaluation Standards: A Unified Framework for Evaluating the Korean Capabilities of Language Models

Hanwool Lee*, Dasol Choi*, Sooyong Kim, Ilgyun Jung, Sangwon Baek, Guijin Son, Inseon Hwang, Naeun Lee, Seunghyeok Hong

LREC 2026

Paper

Pushing on Multilingual Reasoning Models with Language-Mixed Chain-of-Thought

Guijin Son, Donghun Yang, Hitesh Laxmichand Patel, Amit Agarwal, Hyunwoo Ko, Chanuk Lim, Srikant Panda, Minhyuk Kim, Nikunj Drolia, Dasol Choi, Kyong-Ha Lee, Youngjae Yu

ICLR 2026

Paper

Evaluation Framework of Large Language Models in Medical Documentation: Development and Usability Study

Junhyuk Seo*, Dasol Choi*, Wonchul Cha, Haanju Yoo, Namkee Oh, Yongjin Yi, Gyehwa Lee, Edward Choi, Taerim Kim

JMIR 2024

Paper

XL-SafetyBench: A Country-Grounded Cross-Cultural Benchmark for LLM Safety and Cultural Sensitivity

Dasol Choi, Eugenia Kim, Jaewon Noh, Sang Seo, Eunmi Kim, Myunggyo Oh, Yunjin Park, Brigitta Jesica Kartono, Josef Pichlmeier, Helena Berndt, Sai Krishna Mendu, Glenn Johannes Tungka, Özlem Gökçe, Suresh Gehlot, Katherine Pratt, Amanda Minnich, Haon Park

Under review

Paper

When Context Flips, Safety Breaks: Diagnosing Brittle Safety in Aligned Language Models

Dasol Choi*, Alex Kwon*

Under review

Paper

EgoSafetyBench: A Diagnostic Egocentric Video Benchmark for Evaluating Embodied VLMs as Runtime Safety Guards

Siddhant Panpatil, Arth Singh, Mijin Koo, Chaeyun Kim, Haon Park, Dasol Choi†

Under review

Paper

When Good Sounds Go Adversarial: Jailbreaking Audio-Language Models with Benign Inputs

Hiskias Dingeto*, Taeyoun Kwon*, Dasol Choi*, Bodam Kim, DongGeon Lee, Haon Park, JaeHoon Lee, Jongho Shin

Under review

Paper

NOLLI: A Difficulty-Calibrated Puzzle Benchmark for Diagnosing the English-Korean Performance Gap

Dasol Choi*, Joonyong Park*, Daegon Yu, Soo Yong Kim, Youngsook Song, Seunghyeok Hong

Under review

Paper

Service

Judgement Day: AI Red Team ArenaApr. – May 2026

Project Lead & Organizer, AIM Intelligence & Korea AI Safety Institute (AISI)

Planned and ran the competition end-to-end — adversarial scenario design (2 tracks, 8 high-stakes scenarios), multimodal attack channels, and dynamic bounty scoring.

Seoul Forum on AI Safety & Security (SFASS) 2026Jul. 2026

Organizing Member & Speaker

Co-organized the forum program, including the Red-Teaming Workshop and Live Red-Teaming Challenge (with AIM Intelligence).

Talk: “Judgement Day: Insights from Multimodal Red-Teaming Challenge.”

Program Committee Member

AAAI 2026, 2027

Vitæ

Full CV available in PDF.

Dasol Choi
Last updated on June, 2026