Junseo Jang

Junseo Jang

AI Research Engineer

About

Hi, I’m Junseo, an AI Research Engineer at the KT Responsible AI Center. My research interests lie in AI and agent safety, with my recent work focusing on the safety evaluation and alignment of agentic LLMs. Previously, I completed my M.S. at the NLP Lab, Konkuk University, advised by Prof. Harksoo Kim and interned with the HyperCLOVA X Pretraining team at NAVER Cloud.

Publications

* indicates equal contribution

2026
KidnapRAG: A Black-Box Attack for Hijacking Reasoning in Agentic Retrieval-Augmented Generation Systems
Chanwoo Choi, Euntae Kim, Kyuho Lee, Youngsam Chun, Jinhee Jeong, Eunmi Kim, Myunggyo Oh, Junseo Jang, Buru Chang
Preprint Under Review
Paper
2025
Exploring the Impact of Instruction-Tuning on LLM's Susceptibility to Misinformation
Kyubeen Han*, Junseo Jang*, Hongjin Kim, Geunyoung Jung, Harksoo Kim
ACL 2025 (Main)
Paper
Small Changes, Big Impact: How Manipulating a Few Neurons Can Drastically Alter LLM Aggression
Jaewook Lee*, Junseo Jang*, Oh-Woong Kwon, Harksoo Kim
ACL 2025 (Main)
Paper
AESPA: Automated Essay Scoring using Polished Argument Feature Weights
Junseo Jang*, Joosang Lee*, Yejin Lee, Harksoo Kim
Pattern Analysis and Recognition 28, 141 (2025)
Paper
2024
Enhancing the Reasoning Capability of SLM through Bidirectional Chain of Thought Distillation
Junseo Jang, Hongjin Kim, Harksoo Kim
HCLT 2024
Paper
2023
Continuous Korean Named Entity Recognition Using Knowledge Distillation
Junseo Jang, Seongsik Park, Harksoo Kim
HCLT 2023
Paper

Technical Report

HyperCLOVA X 32B Think
NAVER Cloud HyperCLOVA X Team
arXiv preprint, Jan 2026
Paper
HyperCLOVA X THINK Technical Report
NAVER Cloud HyperCLOVA X Team
arXiv preprint, Jun 2025
Paper

Vitæ

KT Corp. Jul 2025 – Present
Responsible AI Center · Full Time
  • AI Research Engineer at the Responsible AI Center, focusing on AI & Agent Safety.
  • Established AI safety governance and conducted safety & impact assessments for AI models and Agents
  • Researched safety alignment methods for tool-calling language models.
NAVER Cloud May – Jul 2025
HyperCLOVA Team · Internship
  • Research Intern on the HyperCLOVA X Pre-training Team.
  • Developed a code data synthesis and quality validation pipeline for LLM pre-training.
Konkuk University Sep 2023 – Aug 2025
M.S. in Artificial Intelligence
  • Advisor: Harksoo Kim · GPA: 4.21/4.5
AgileSoda Jul – Dec 2022
NLP Team · Internship
  • Continual learning for medical NER
  • ICD code mapping with XAI
Konkuk University Mar 2017 – Aug 2023
B.S. in Computer Science
  • GPA: 4.1/4.5