Seongtae Hong
Hi! I’m Seongtae Hong, a Ph.D. student at the NLP & AI Lab, Korea University, advised by Prof. Heuiseok Lim. Previously, I worked as a researcher at the Artificial Intelligence Research Center of the Korea Electronics Technology Institute (KETI). I received my B.S. from Changwon National University, where I was a member of the AIR Lab. My research focuses on multilingualism and information retrieval, with broader interests in retrieval-augmented generation (RAG) systems, spanning retrieval, reranking, and generation.
Seoul, Republic of Korea
News
- Our paper MLAIRE is accepted to NeurIPS 2026 E&D!
- Our paper Skin-Deep is accepted to AACL-IJCNLP 2026 Findings!
- 4 papers accepted to EMNLP 2026: 2 Main (1 oral) and 2 Findings.
- Our paper on trainable prefix tokens for retrieval embeddings is accepted to CIKM 2026!
- We released LAMAR-600m, our multilingual reranker, with its paper.
Research
My research is driven by the goal of making language technologies and knowledge more accessible across languages, communities, and resource constraints. I explore how models can transfer and generalize capabilities across languages and tasks through cross-lingual alignment, representation learning, and efficient adaptation. Ultimately, I aim to reduce dependence on language-specific resources and task-specific solutions, making language technologies more reliable and adaptable to diverse real-world needs.
How can we find knowledge beyond the language of a query?
I study multilingual representations and sparse and dense retrieval, with an interest in transferring retrieval capabilities across languages and resource settings.
The next question: which of the retrieved passages should a system trust and use?
Publications
2026
* denotes equal contribution · † denotes corresponding author.
MLAIRE: Multilingual Language-Aware Information Retrieval Evaluation Protocol
Youngjoon Jang, Seongtae Hong, Hyeonseok Moon, Heuiseok Lim†
Reducing Representational Interference in Retrieval Embeddings with Trainable Prefix Tokens
Seongtae Hong, Youngjoon Jang, Jungseob Lee, Chanjun Park†, Heuiseok Lim†
Skin-Deep: A Geometric Diagnostic for Alignment Fragility in Large Language Model Representations
Dongyub Jude Lee, Jungseob Lee, Seungyoon Lee, Seongtae Hong, Suhyune Son, Sugyeong Eo, Jaehyung Seo, Heuiseok Lim†
Distilling Directional Verification
Jungseob Lee, Sugyeong Eo, Seongtae Hong, Seungyoon Lee, Chanjun Park, Jaehyung Seo, Heuiseok Lim†
Refusal Localizes, the Damage Relocates: Safety Layers Under Few-Sample Fine-Tuning
Jungseob Lee, Dongyub Jude Lee, Sugyeong Eo, Seongtae Hong, Seungyoon Lee, Heuiseok Lim†
DART: Draft-Agreement Routing for Training-Free Adaptive Thinking Budgets in Hybrid Reasoning Models
Jungseob Lee, Seongtae Hong, Seungjun Lee, Jaehyung Seo, Junyoung Son, Sugyeong Eo, Chanjun Park, Hyeongju Park, Hyeonseok Moon, Heuiseok Lim†
SHIFT: Semantic Harmonization via Index-side Feature Transformation for Multilingual Information Retrieval
Youngjoon Jang, Seongtae Hong, Hyeonseok Moon, Heuiseok Lim†
MIMO: Multilingual Information Retrieval via Monolingual Objectives
Youngjoon Jang, Seongtae Hong, Heuiseok Lim†
SemBridge: Language Transfer in Sparse Encoders via Multilingual Semantic Bridges
Seongtae Hong, Youngjoon Jang, Jia-Heui Ju, Hyeonseok Moon, Heuiseok Lim†
RenderRank: Learning to Rerank Text with Compressed Visual Tokens
Seongtae Hong, Youngjoon Jang, Jungseob Lee, Hyeonseok Moon, Heuiseok Lim†
Where Activation Sparsity and KV-Cache Sparsity Cross in LLM Decoding
Jungseob Lee, Seungyoon Lee, Seongtae Hong, Sugyeong Eo, Heuiseok Lim†
Vision Is Not Overhead: One-Pass Block Drafting for Lossless Speculative Decoding in Vision-Language Models
Jungseob Lee, Seongtae Hong, Dongyub Jude Lee, Chanjun Park, Jaehyung Seo, Sugyeong Eo, Heuiseok Lim†
Funnel of Thoughts: Efficient Test-Time Scaling via Early Voting and Rollout Pruning
Chanhee Park, Sungbin Han, Jeongho Yoon, Seongtae Hong, Heuiseok Lim†
LAMAR: An Open Language-Aware Multilingual Alignment Reranker
Seongtae Hong, Youngjoon Jang, Jungseob Lee, Seungyoon Lee, Heuiseok Lim†
Improving korean-english cross-lingual retrieval: A data-centric study of language composition and model merging
Youngjoon Jang, Junyoung Son, Taemin Lee, Seongtae Hong, Hyeonseok Moon, Seungyoon Lee, Andrew Matteson, Heuiseok Lim†
Beyond Hard Negatives: The Importance of Score Distribution in Knowledge Distillation
Youngjoon Jang, Seongtae Hong, Hyeonseok Moon, Heuiseok Lim†
QUERY-SYNERGY: Leveraging High-Resource Languages for Improving Retrieval Performance Across Multiple Languages
Seongtae Hong, Jungseob Lee, Hyeonseok Moon, Seungyoon Lee, Youngjoon Jang, Heuiseok Lim†
I Don’t Need Solution. I Need Emotional Support: Empathetic LLMs based on Emotional Validation
Suhyune Son, Jungwoo Lim, Myunghoon Kang, Seongtae Hong, Yuna Hur, Evelyn H Zi, Heuiseok Lim†
CLEAR: Cross-Lingual Enhancement in Retrieval via Reverse-training
Seungyoon Lee, Minhyuk Kim, Seongtae Hong, Youngjoon Jang, Dongsuk Oh, Heuiseok Lim†
Beyond Penalizing Mistakes: Stabilizing Efficiency Training in Large Reasoning Models via Adaptive Correct-Only Rewards
Jungseob Lee, Seungyoon Lee, Seongtae Hong, Minhyuk Kim, Chanjun Park, Heuiseok Lim†
Rescaling MLM-Head for Neural Sparse Retrieval
Youngjoon Jang, Seongtae Hong, Jonah Turner, Heuiseok Lim†
Improving semantic proximity in english-centric information retrieval through cross-lingual alignment
Seongtae Hong, Youngjoon Jang, Jungseob Lee, Hyeonseok Moon†, Heuiseok Lim†
Semantic Fragment Similarity Representation Learning for Information Retrieval
Seongtae Hong, Hyeonseok Moon, Youngjoon Jang, Junyoung Son, Jaehyung Seo, Heuiseok Lim†
2025
Metric Calculating Benchmark: Code-Verifiable Complicate Instruction Following Benchmark for Large Language Models
Hyeonseok Moon, Seongtae Hong, Jaehyung Seo, Heuiseok Lim†
Cross-lingual optimization for language transfer in large language models
Jungseob Lee*, Seongtae Hong*, Hyeonseok Moon, Heuiseok Lim†
Revise: A framework for revising ocred text in practical information systems with data contamination strategy
Gyuho Shim*, Seongtae Hong*, Heuiseok Lim†
From Ambiguity to Accuracy: The Transformative Effect of Coreference Resolution on Retrieval-Augmented Generation systems
Youngjoon Jang*, Seongtae Hong*, Junyoung Son, Sungjin Park, Chanjun Park, Heuiseok Lim†
Semantic Aware Linear Transfer by Recycling Pre-trained Language Models for Cross-lingual Transfer
Seungyoon Lee, Seongtae Hong, Hyeonseok Moon, Heuiseok Lim†
MIGRATE: Cross-Lingual Adaptation of Domain-Specific LLMs through Code-Switching and Embedding Transfer
Seongtae Hong, Seungyoon Lee, Hyeonseok Moon, Heuiseok Lim†
2024
Intelligent predictive maintenance RAG framework for power plants: Enhancing QA with StyleDFS and domain specific instruction tuning
Seongtae Hong*, Joong Min Shin*, Jaehyung Seo, Taemin Lee, Jeongbae Park, Cho Man Young, Byeongho Choi, Heuiseok Lim†
Translation of Multifaceted Data without Re-Training of Machine Translation Systems
Hyeonseok Moon, Seungyoon Lee, Seongtae Hong, Seungjun Lee, Chanjun Park, Heuiseok Lim†
Kocommongen v2: A benchmark for navigating korean commonsense reasoning challenges in large language models
Jaehyung Seo, Jaewook Lee, Chanjun Park, SeongTae Hong, Seungjun Lee, Heuiseok Lim†
Models
View all on Hugging Facenlpai-lab/LAMAR-600m
LAMAR-600m
A language-aware multilingual cross-encoder reranker trained for semantic relevance and query–document language coherence.
600M · 51 languages · MIT
nlpai-lab/RenderRank-2B
RenderRank-2B
A pairwise document reranker that renders text as compact visual tokens to handle more content within a fixed context budget.
2B · English · Apache-2.0
nlpai-lab/MixLID
MixLID
A token-level language identification model for monolingual and mixed-language text across 111 language labels.
mmBERT-small · 111 labels · MIT
nlpai-lab/KURE-Reranker-nano
KURE-Reranker-nano
A lightweight Korean–English cross-encoder reranker built for strong retrieval quality with efficient inference.
149M · Korean · English · Apache-2.0
Academic Activities
Conference Reviewer
- 2026: NeurIPS, SIGIR, ACL, EMNLP, AAAI, ICLR
- 2025: ACL, EMNLP