Publications

Publications

List of Publications

International Publications

  1. Distilling Long-CoT Reasoning through Collaborative Step-wise Multi-Teacher Decoding
    Taewon Yun, Jisu Shin, Seunghwan Bang, Jeonghwan Choi, Hwanjun Song
    (Under Review) ARR Oct

  2. Completing Missing Annotation: Multi-Agent Debate for Accurate and Scalable Relevant Assessment for IR Benchmarks
    Minjeong Ban*, Jeonghwan Choi *, Hyangsuk Min*, Nicole Hee-Yeon Kim, Minseok Kim, Jae-Gil Lee, Hwanjun Song (with Meta)
    (Under Review) International Conference on Learning Representations (ICLR), 2026
    [PDF]

  3. BRIDGE: Toward Faithful RAG Benchmarking via Retrieval-Generation Alignment
    Minjeong Ban*, Jeonghwan Choi *, Hyangsuk Min*, Heeyeon Kim, Seunghwan Bang, and Hwanjun Song
    ACM International Conference on Information and Knowledge Management (RDGENAI@CIKM), 2026

  4. Ext2Gen: Alignment through Unified Extraction and Generation for Robust Retrieval-Augmented Generation
    Hwanjun Song, Jeonghwan Choi, Minseok Kim
    ACM International Conference on Web Search and Data Mining (WSDM), 2026
    [PDF]

  5. Word2Passage: Word-level Importance Re-weighting for Query Expansion
    Jeonghwan Choi, Minjeong Ban, Minseok Kim, Hwanjun Song (with Meta)
    Findings of the Association for Computational Linguistics (Findings of ACL), 2025
    [PDF]

  6. Learning to Verify Summary Facts with Fine-Grained LLM Feedback
    Jihwan Oh, Jeonghwan Choi, Nicole Hee-Yeon Kim, Taewon Yun, Hwanjun Song
    International Conference on Computational Linguistics (COLING), 2025
    [PDF]

Domestic Publications

  1. A Query Decompositional Framework for Fine-grained RAG Assessment
    Jeonghwan Choi, Minjeong Ban, Nicole Hee-Yeon Kim, Taewon Yun, Yuho Lee, Hwanjun Song
    (Under Review) Korea Software Congress (KSC), 2026

  2. A Benchmark Dataset for Retrieval-Augmented Generation Considering Performance Alignment between Retrieval and Generation
    Minjeong Ban*, Jeonghwan Choi *, Hyangsuk Min*, Taewon Yun, Jihwan Oh, Hwanjun Song
    Korea Computer Congress (KCC), 2025

  3. Robust Dataset Condensation via Semi-Supervised Learning
    Heeyeon Kim, Jeonghwan Choi, Yuho Lee, Hwanjun Song
    Korea Computer Congress (KCC), 2025

  4. Improving Language Model Quality through LLM-based Fine-Grained Hallucinated Summary Generation
    Jihwan Oh, Jeonghwan Choi, Nicole Hee-Yeon Kim, Taewon Yun, Hwanjun Song
    KIISE Transactions on Computing Practices, (KTCP), 2025

  5. Training Summary Evaluation Model with Fine-Grained LLM Feedback
    Jihwan Oh, Jeonghwan Choi, Taewon Yun, Hyangsuk Min, Hwanjun Song
    Korea Software Congress (KSC), 2024

  6. Improving the Text Summary Quality Through Understanding the Hallucination Level of Summarization Using Large Language Models
    Jihwan Oh, Jeonghwan Choi, Nicole Hee-Yeon Kim, Hwanjun Song
    Korea Computer Congress (KCC), 2024
    [PDF]

List of Participated Projects

National Research Foundation of Korea (NRF)

  • Research Team Member - Language Model Reliability Enhancement System, Apr 2024 - Mar 2027 (Seoul, South Korea)

    • Developing automated evaluation system for reliable assessment of AI-generated text using high-quality data and automated quality assessment

    • Contributing to methodologies that reduce hallucinations and biases in generative language models while enhancing completeness and conciseness

    • Working on benchmark dataset construction for AI model reliability assessment

Institute of Information & Communications Technology Planning & Evaluation (IITP)

  • Research Team Member - AI Model Reliability Through Domain-Specific Assessment, Sep 2024 - Aug 2027 (Seoul, South Korea)

    • Enhancing AI model reliability through Korea-Canada international joint research collaboration (KAIST-GIST-UBC)
    • Constructing expert-reviewed, high-quality, value-aligned benchmark datasets specialized in language, mathematics, and medicine
    • Developing core technologies for high-performance automated value alignment evaluation

Korea Institute of Science and Technology Information (KISTI)

  • Project Manager - Scientific Information RAG System Quality Improvement, Jun 2025 - Dec 2025 (Daejeon, South Korea)
    • Improving output quality of scientific information specialized RAG (Retrieval-Augmented Generation) system
    • Identifying unique requirements of domain-specific RAG for scientific and technical information
    • Developing optimized retrieval strategies, generation techniques, and evaluation metrics for science domain RAG systems

Korea Institute of Science and Technology Information (KISTI)

  • Project Manager - RAG-Supported Data Building for Scientific LLMs Jun 2024 - Dec 2024 (Daejeon, South Korea)
    • Developed Korean science domain-specialized RAG (Retrieval-Augmented Generation) system to enhance accuracy and reliability of scientific information utilization
    • Contributed to data construction algorithms for improving LLM trustworthiness in science and technology domains