Chaeyun Kim

About Me

I obtained my M.S. in Data Science from Seoul National University's Visual Information Processing Lab, advised by Prof. Joonseok Lee.

Long-term Goal: Transforming multimodal LLMs into systems that are controllable, interpretable, and safe — mechanisms that generalize across modalities, languages, and deployment contexts.

Research Interests

Interpretability, Robustness, and Alignment
Reasoning & Grounding: Spatial-geometric reasoning and multimodality
AI Safety: Jailbreak methods, AI Guardrail, AI Red-teaming

News

[Apr. 2026] One paper about strategy-response multiplex network for automated LLM red-teaming accepted to ACL 2026 Findings.
[Feb. 2026] One paper about multimodal object hallucination evaluation accepted to CVPR 2026.
[Jan. 2026] One paper about culturally adaptive red-teaming accepted to ICLR 2026.
[Jan. 2026] One paper about motion-aware referring image segmentation accepted to AISTATS 2026.
[Mar. 2025] Started working at Selectstar as AI Researcher.
[Jul. 2024] One paper about referring image segmentation accepted to ECCV 2024.
[Sep. 2022] Started M.S. in Data Science at Seoul National University.

Publications

In Review

FinRED: An Expert-Guided Red-Teaming Benchmark for Financial LLM Safety

Chaeyun Kim*, Dae-Young Park*, Junghwan Kim, Jinyoung Jeong, Eunji Song, YongTaek Lim, Minwoo Kim

In Review (KDD Datasets and Benchmark Track), 2026.

Under Review

ACL 2026 Findings

STAR-Teaming: A Strategy-Response Multiplex Network Approach to Automated LLM Red Teaming

MinJae Jung, YongTaek Lim, Chaeyun Kim, Junghwan Kim, Kihyun Kim, Minwoo Kim

Findings of the Association for Computational Linguistics (ACL Findings), 2026.

Coming Soon

CVPR 2026

Fine-Grained Multi-Image Object Hallucination Benchmark

Joonki Min*, Chaeyun Kim*, Hyungwook Choi, Yejin Kim, Kihyun Kim, Yohan Jo, Joonseok Lee

IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), 2026.

Coming Soon

ICLR 2026

CAGE: A Framework for Culturally Adaptive Red-Teaming Benchmark Generation

Chaeyun Kim, Yongtaek Lim, Kihyun Kim, Junghwan Kim, Minwoo Kim

International Conference on Learning Representations (ICLR), 2026.

PDF Code

AISTATS 2026

Towards Motion-aware Referring Image Segmentation

Chaeyun Kim*, Seunghoon Yi*, Yejin Kim, Yohan Jo, Joonseok Lee

International Conference on Artificial Intelligence and Statistics (AISTATS), 2026.

PDF Code Project Page

ECCV 2024

Finding NeMo: Negative-mined Mosaic Augmentation for Referring Image Segmentation

Seongsu Ha*, Chaeyun Kim*, Donghwa Kim*, Junho Lee, Sangho Lee, Joonseok Lee

European Conference on Computer Vision (ECCV), 2024.

PDF Code Project Page

Experience

Education

M.S., Data Science (Visual Information Processing Lab), Seoul National University, Feb. 2025
B.S., International Relations, Seoul National University, Feb. 2022
Exchange Student, University of Washington, Seattle, Jan. 2020 – Jul. 2020

Chaeyun Kim

About Me

Research Interests

News

Publications

Experience

Education

Work Experience

Services

Conference Reviewers