Institution profile

Chung-Ang University

Academic institutionasia · kr
Official website
Research library301linked papers
Opportunities0open roles
Selected work

Representative Papers

SummPilot: Bridging Efficiency and Customization for Interactive Summarization System

Apr 11, 2025AAAI Conference on Artificial Intelligence

This work proposes SummPilot, an interactive and customizable summarization system powered by large language models that reconciles the efficiency of automated summarization with users’ personalized needs. For the first time, it integrates an interactive semantic graph, entity clustering, and an interpretable evaluation mechanism within a unified framework, enabling users to actively shape summary generation through visual exploration of semantic structures and key entities. By balancing automation with user control, SummPilot significantly enhances the utility and adaptability of generated summaries, as demonstrated in user studies. The results validate its effectiveness across diverse application scenarios requiring tailored summarization.

1 citationsRead paper

Questioning Internal Knowledge Structure of Large Language Models Through the Lens of the Olympic Games

Sep 10, 2024arXiv.org

This study identifies a structural deficiency in large language models (LLMs) regarding historical Olympic medal knowledge: while LLMs accurately retrieve national medal counts (Task 1, >90% accuracy), they exhibit severe limitations in ranking logic reasoning (Task 2, <35% accuracy). We construct a systematic, fine-grained Olympic medal dataset and evaluate multiple state-of-the-art LLMs via zero-shot prompting. Our empirical analysis reveals— for the first time—that LLMs’ knowledge representation is biased toward factual recall rather than relational reasoning: they reliably encode “how many” but fail to consistently infer “which rank.” This finding indicates a fundamental divergence between LLMs’ internal knowledge organization and human-like structured reasoning, challenging the implicit assumption that LLMs serve as general-purpose reasoning engines. To support reproducible evaluation of structured reasoning capabilities, we publicly release all code, data, and model outputs, establishing a benchmark for assessing relational inference in foundation models.

1 citationsRead paper
Recent publications

Latest Papers