Jueun Kim

I am a Ph.D. student at Data Systems Lab @ Pohang University of Science and Technology (POSTECH), under the supervision of Professor Wook-Shin Han.

My research is on multi-hop, multi-modal information retrieval: finding evidence that is scattered across text, tables, and other modalities, where no single query retrieves everything a task needs.

More broadly, I am interested in what retrieval should become when its consumer is an agent carrying out a task rather than a person reading through results. An agent needs the right evidence at the right step of its work, and has to decide for itself what to look for next. Agent memory connects naturally here — what an agent should keep from its own experience, and how that memory should be organized so it stays useful over long horizons instead of accumulating into noise.

news

Sep 28, 2026	🔥 Our Paper accepted at NeurIPS 2026: “Hi-Q: Hierarchical Evidence-guided Query Refinement for Multi-Hop Question Answering”
Jan 26, 2026	🔥 Our Paper accepted at ICLR 2026: “SPARTA: Scalable and Principled Benchmark of Tree-Structured Multi-hop QA over Text and Tables”
Nov 01, 2025	🏅 Received the POSTECHIAN Fellowship (Creative).
Sep 01, 2025	🚀 Started Ph.D. program in Computer Science and Engineering at POSTECH.
Aug 08, 2025	🎓 Graduated from POSTECH with a B.S. degree in Computer Science and Engineering.

selected publications

CANOPY: Adaptive-Granularity Evidence Compression for Multimodal RAG

Hyojeong Yun^*, Jueun Kim^*, Wook-Shin Han (^*equal contribution)

arXiv preprint arXiv:2610.00923, 2026

Abs arXiv Website

Multimodal RAG retrieves text, tables, images, and videos, but choosing a retrieval granularity does not determine how much context to retain within each item. We introduce CANOPY, a framework for adaptive-granularity post-retrieval evidence compression that represents retrieved items as hierarchies and selects multiple regions at different granularities, requesting targeted follow-up retrieval when the accumulated evidence is judged insufficient.
Hi-Q: Hierarchical Evidence-guided Query Refinement for Multi-Hop Question Answering

Jueun Kim, Sungho Park, Wook-Shin Han

The Fortieth Conference on Neural Information Processing Systems (NeurIPS), 2026

Abs arXiv Website

A central bottleneck in multi-hop QA is that the granularity at which a question is expressed differs from the granularity at which corpus evidence is retrievable. We formulate this as retrievable granularity discovery and introduce Hi-Q, an evidence-conditioned framework for hierarchical query refinement that grows a query tree whose topology is determined by corpus support signals rather than by a fixed decomposition template or a pre-built graph.
SPARTA: Scalable and Principled Benchmark of Tree-Structured Multi-hop QA over Text and Tables

Sungho Park, Jueun Kim, Wook-Shin Han

The Fourteenth International Conference on Learning Representations(ICLR), 2026

Abs arXiv Website

SPARTA is a fully automated SQL-centric pipeline that constructs a tree-structured multi-hop QA benchmark by unifying structured and unstructured evidence from tables and text into a single relational representation.