Research

Any research topic related to natural language processing and A(G)I  is within the interest of HYU NLP Lab. members.
Specifically, we are recently paying more attention to the topics such as:

• Reliable AGI

  • Safety and alignment

  • Abstention & refusal with epistemic humility

  • Uncertainty

• Science of LLMs: Studying LLMs as an object of science

  • Probing LLMs from a linguistic (or other disciplinary) perspective

  • Mechanistic intepretability

• Effective utilization of LLMs

  • In-context learning and steering

  • Knowledge managment: unlearning, knowledge editing / conflict, knowlege graph

  • Dialogue systems with long-term memory

  • Scaling test-time compute (meta-generation): RL-based methods

• LLM agents

  • Multi-agents

  • Simulated / emboddied agents

  • Self-improvement, self-refine, self-reflection

• Equivalent benefits of NLP/LLMs

  • • Multilingual / multicultural language models

    • Korean NLP

    • Application to other fields: AI for Social Science

• Embedding and Retrieval

  • • Sentence representation learning (contrastive learning)

    • Retrieval-augmented generation

• Multimodality

  • • Multi-modal (or cross-modal) multi-hop reasoning

    • Multimodal retrieval

• 책임 연구

  • • 초장기기억 기반 대화형 에이전트, 한국연구재단 (NRF)-우수신진연구, 2025-2028

    • 음성인식 고도화를 위한 복합의도 모델 연구, 현대자동차 (현대 NGV), 2024-2026

    • 자연어 기반 복합 의도 알고리즘 연구, 현대자동차 (현대 NGV), 2023-2024

    • 대규모 언어 모델 기반 고품질 문장 표현 학습을 위한 효율적인 방법론 분석 및 개발, 한국연구재단 (NRF)-기본연구, 2022-2024

    • 음성인식 자연어처리엔진의 차량 도메인 언어 이해 향상을 위한 사전학습언어모델 아키텍처 고도화, 현대자동차 (현대NGV), 2022-2023

• 공동 연구

  • • 인공지능반도체대학원(한양대학교), 정보통신기획평가원 (IITP), 2023-

    • 초연결사회 위험 관리를 위한 빅데이터 기반 사회 환경 실시간 모니터링/사회 시뮬레이션 시스템, 한국연구재단 (NRF), 2022-2025

    • 한양대학교 컴퓨터소프트웨어학과 BK21-FOUR 인공지능 혁신인재 교육단, 한국연구재단 (NRF), 2022-

    • 한양대학교 인공지능대학원지원사업, 정보통신기획평가원 (IITP), 2021-

• Visit here!

(24/08/22) Yejin and Jii have graduated with their Master's degrees. Wish them all the best! Yejin is going to begin her Ph.D. studies!

(24/06/14) Four (3 oral and 1 poster) papers have been accepted for presentation at KCC 2024. Congrats to Yejin, Jungyeon, Jisu, Youngwoo, Kang Min, Dong Geon, and Jungmin! Also excited to share 한국어 발화의 다중 의도 감지 연구 (Multi-Intent Detection for Korean Spoken Language) has received the Outstanding Paper Award (우수논문상) from the conference!

(24/05/16) Two papers, 1. Analysis of Multi-Source Language Training in Cross-Lingual Transfer (authors: Seong Hoon, Taejun, Jinhyeon) 2. Hyper-CL: Conditioning Sentence Representations with Hypernetworks (authors: Young Hyun, Jii, Changhyeon), have been accepted for presentation at ACL 2024. Big congrats to the authors!

(24/02/20) One paper (BlendX: Complex Multi-Intent Detection with Blended Patterns) has been accepted for presentation at LREC-COLING 2024. Big congrats to Yejin, Jungyeon, and Kangsan!

(24/02/15) Jinhyeon, Young Hyun, Taejun, and Seong Hoon have graduated with their Master's degrees. Wish them all their best!

(23/11/22) One paper has been accepted to KSC 2023. Congrats to Jii!

(23/10/08) Two papers (X-SNS: Cross-Lingual Transfer Prediction through Sub-Network Similarity & Universal Domain Adaptation for Robust Handling of Distributional Shifts in NLP) have been accepted at EMNLP 2023 (Findings).  X-SNS is especially our first internal project that will be showcased at a major conference.  Congrats to Taejun, Jinhyeon, Deokyoung, and Seong Hoon. See you in Singapore!
(23/09/25) Two papers have been accepted to HCLT 2023. Congrats to Jinhyeon and Taejun!

(23/06/30) Updated the contents regarding our journey to KCC 2023, which took place on the Jeju island. At the conference, Seong Hoon presented his work titled Enhanced Zero-Shot Cross-Lingual Transfer with the Diversification of Source Languages.

(22/11/25) One paper (Prompt-Augmented Linear Probing: Scaling Beyond The Limit of Few-shot In-Context Learners) has been accepted to AAAI 2023. Congrats!

(22/10/16) Taeuk, Kangsan, Jinhyeon, Minjin, and Kyumin participated in COLING 2022.

(22/10/07) Two papers (Ground-Truth Labels Matter: A Deeper Look into Input-Label Demonstrations and Enhancing Out-of-Distribution Detection in Natural Language Understanding via Implicit Layer Ensemble) for which Taeuk participated as a corresponding author have been accepted to EMNLP 2022 (1 Main, 1 Findings). They are the results of a collaboration with SNU and Naver. See you in Abu Dhabi!

(22/10/04) One paper (Comparison and Analysis of Unsupervised Contrastive Learning Approaches for Korean Sentence Representations) has been accepted to HCLT 2022 (authors: Young Hyun, Kyumin, Minjin, Jii, Kangsan, and Taeuk). Congrats!

(22/08/17) Revisiting the Practical Effectiveness of Constituency Parse Extraction from Pre-trained Language Models (authors: Taeuk) has been accepted to COLING 2022. See you soon in Gyeongju!

(22/05/22) Taeuk has been invited to serve as an Area Chair for the Unsupervised and Weakly-supervised Methods Track of EMNLP 2022.

(22/04/13) HYU at SemEval-2022 Task 2: Effective Idiomaticity Detection with Consideration at Different Levels of Contextualization (authors: Youngju and Taeuk) has been accepted to SemEval 2022. Congrats!

(21/10/15) Taeuk has been nominated to serve as an Action Editor for the ACL Rolling Review (a new reviewing platform for *ACL conferences)!

(21/08/26) One paper (Multilingual Chart-based Constituency Parse Extraction from Pre-trained Language Models) is accepted at EMNLP 2021 Findings.

(21/08/03) Taeuk presented his work Self-Guided Contrastive Learning for BERT Sentence Representations at ACL 2021.