<aside> 📩 [email protected]
</aside>
<aside> <img src="data:image/png;base64,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" alt="data:image/png;base64,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" width="40px" /> @yeonjun-in
</aside>
<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/36c7f341-a52a-40e5-a549-3fb4133e27fb/46a01e87-8cf8-4ca8-98b3-f6bc18fbef39/linkedin_480px.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/36c7f341-a52a-40e5-a549-3fb4133e27fb/46a01e87-8cf8-4ca8-98b3-f6bc18fbef39/linkedin_480px.png" width="40px" /> @Yeonjun In
</aside>
<aside> <img src="https://prod-files-secure.s3.us-west-2.amazonaws.com/36c7f341-a52a-40e5-a549-3fb4133e27fb/f30304aa-3449-4968-94ed-4711f30b0b92/icons8-twitter-48.png" alt="https://prod-files-secure.s3.us-west-2.amazonaws.com/36c7f341-a52a-40e5-a549-3fb4133e27fb/f30304aa-3449-4968-94ed-4711f30b0b92/icons8-twitter-48.png" width="40px" /> @yeonjun_in
</aside>
I’m Yeonjun In, a Ph.D. candidate in the Industrial and Systems Department at KAIST, where I am fortunate to be advised by Prof. Chanyoung Park. I’m actively on research with my best colleagues at Data Science and Artificial Intelligence Lab :ori:.
I work on building reliable and trustworthy AI for diverse real-world applications, including unstructured and structured data. (1) Model Robustness against Data-Centric Issues In real-world applications, diverse data-centric issues hinder the training of robust AI models. My research addresses this challenge by developing: (a) robust representation learning algorithms, and (b) anomaly detection algorithms.
(2) Harmfulness and Safety of LLMs As the use of LLMs continues to grow, their associated safety risks and harmful use cases have become increasingly evident. My research addresses this challenge by (a) uncovering hidden safety risks in current LLMs and (b) mitigating these risks through LLM alignment.
✔️ New preprints out: Personalize-then-Store and PAIR. ✔️ A paper got accepted at ICML 2026 Workshop on Failure Modes of Agentic AI ✔️ Two papers got accepted at ICML 2026. ✔️ A paper got accepted at ACL 2026 (Main). ✔️ A paper got accepted at NeurIPS 2025 Workshop: Reliable ML from Unreliable Data. ✔️ A paper got accepted at NeurIPS 2025.
Full publication list: Google Scholar
(*: joint first author)
PAIR: Prefix-Aware Internal Reward Model for Multi-Turn Agent Optimization Wonjoong Kim, Yeonjun In, Sangwu Park, Dongha Lee, Chanyoung Park Preprint. [paper][code] Collaboration with Yonsei University
Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-Horizon Agents Yeonjun In, Wonjoong Kim, Sangwu Park, Kanghoon Yoon, Chanyoung Park KDD 2026 Workshop on Personal Intelligence in Agentic AI Era (Oral). [paper][code] 📰 Must-read research of the week by Turing Post
Rethinking Failure Attribution in Multi-Agent Systems: A Multi-Perspective Benchmark and Evaluation Yeonjun In, Md Mehrab Tanjim, Jayakumar Subramanian, Sungchul Kim, Uttaran Bhattacharya, Wonjoong Kim, Sangwu Park, Somdeb Sarkhel, Chanyoung Park KDD 2026 Workshop on Agentic Software Engineering (Oral) & ICML 2026 Workshop on Failure Modes of Agentic AI. [paper][dataset][code] Work done during internship at Adobe Research
Beyond the Final Answer: Evaluating the Reasoning Trajectories of Tool-Augmented Agents Wonjoong Kim, Sangwu Park, Yeonjun In, Sein Kim, Dongha Lee, Chanyoung Park ICML 2026. [paper] Collaboration with Yonsei University
