LLM Researcher • AI Agents

✋ Welcome to the Introduction Page of Taeyun Roh

I am Taeyun Roh, an LLM researcher at Korea University DMIS Lab advised by Prof. Jaewoo Kang. My research aims to develop AI agents that better understand the world through effective memory organization and visually grounded reasoning.

  • Focus AI Agents
  • Field Large Language Models
  • Base Seoul, Korea

Research Focus

Visual Memory Systems for AI Agents

I study AI agents that can organize useful memories, understand visual evidence, and improve through post-training. These three directions share one goal: building agents that remember what matters and reason reliably before they act.

01 · Memory Systems

How can agents organize and reuse experience?

CLAG (ACL 2026 Findings) augments AI agents with self-organizing memory, structuring past experiences for more focused retrieval and reuse.

02 · Visual Understanding

How can agents reason from what they see?

SCICON (Preprint, 2026) uses contrastive decoding to ground answers in visual evidence without additional training. Breaking Failure Cascades (Preprint, 2026) extends this direction to medical multimodal reasoning.

03 · Post-Training for Agents

How can agents learn to plan and use tools?

Distilling Expert-level Planning (ICML Workshop 2026) uses reinforcement learning to improve tool use for real-world diabetes prescribing. In Breaking Failure Cascades (Preprint, 2026), step-aware reinforcement learning is tailored to multimodal reasoning failures.

Overview

Quick look at recent updates and research outputs

For a concise snapshot of ongoing work, check recent lab and project updates in News, and browse full papers, preprints, and publication records in Publications.