Minkyu Kim

Minkyu Kim

AI Researcher @ KRAFTON

About

I am an AI Researcher at KRAFTON. I like training foundation models, and I spend most of my time on what they should learn from and how to measure what they still cannot do. My goal is models that can see, listen, speak, and act as naturally as people do.

Previously, I did my M.S. at the Graduate School of AI, KAIST, advised by Prof. Jinwoo Shin.

News

  • 🔊 Our team at KRAFTON released Raon, a family of four multimodal models for speech understanding and generation. I worked on the data behind Raon-Speech-9B and Raon-SpeechChat-9B.
  • 📝 1 paper accepted to CVPR 2026 (See and Fix the Flaws).
  • 📝 2 papers accepted to ICLR 2026 (Orak, VLM-SubtleBench).

Publications

* denotes equal contribution · [C] conference paper · [T] technical report
[T1]
Raon-Speech Technical Report
KRAFTON
Technical Report, 2026
📄 Paper 🤗 Models
[C9]
See and Fix the Flaws: Enabling VLMs and Diffusion Models to Comprehend Visual Artifacts via Agentic Data Synthesis
Jaehyun Park, Minyoung Ahn, Minkyu Kim, Jonghyun Lee, Jae-Gil Lee, Dongmin Park
CVPR 2026
📄 Paper
[C8]
VLM-SubtleBench: How Far Are VLMs from Human-Level Subtle Comparative Reasoning?
Minkyu Kim*, Sangheon Lee*, Dongmin Park
ICLR 2026
📄 Paper 💻 Code 🤗 Dataset
[C7]
Orak: A Foundational Benchmark for Training and Evaluating LLM Agents on Diverse Video Games
Dongmin Park*, Minkyu Kim*, Beongjun Choi*, Junhyuck Kim, Keon Lee, Jonghyun Lee, Inkyu Park, Byeong-Uk Lee, Jaeyoung Hwang, Jaewoo Ahn, Ameya S. Mahabaleshwarkar, Bilal Kartal, Pritam Biswas, Yoshi Suhara, Kangwook Lee, Jaewoong Cho
ICLR 2026
📄 Paper 💻 Code
[C6]
Rare-to-Frequent: Unlocking Compositional Generation Power of Diffusion Models on Rare Concepts with LLM Guidance
Dongmin Park, Sebin Kim, Taehong Moon, Minkyu Kim, Kangwook Lee, Jaewoong Cho
ICLR 2025 Spotlight
📄 Paper 💻 Code
[C5]
Test-time Alignment of Diffusion Models without Reward Over-optimization
Sunwoo Kim, Minkyu Kim, Dongmin Park
ICLR 2025 Spotlight
📄 Paper 💻 Code
[C4]
Discovering and Mitigating Visual Biases through Keyword Explanation
Younghyun Kim*, Sangwoo Mo*, Minkyu Kim, Kyungmin Lee, Jaeho Lee, Jinwoo Shin
CVPR 2024 Highlight
📄 Paper
[C3]
Image Clustering Conditioned on Text Criteria
Sehyun Kwon, Jaeseung Park, Minkyu Kim, Jaewoong Cho, Ernest K. Ryu, Kangwook Lee
ICLR 2024
📄 Paper 💻 Code
[C2]
S-CLIP: Semi-supervised Vision-Language Learning using Few Specialist Captions
Sangwoo Mo, Minkyu Kim, Kyungmin Lee, Jinwoo Shin
NeurIPS 2023
📄 Paper
[C1]
SmoothMix: Training Confidence-Calibrated Smoothed Classifiers for Certified Robustness
Jongheon Jeong, Sejun Park, Minkyu Kim, Heung-Chang Lee, Doguk Kim, Jinwoo Shin
NeurIPS 2021
📄 Paper 💻 Code

Education

  • M.S. in Artificial Intelligence, KAIST, Sep 2021 - Aug 2023
  • B.S. in Mathematical Sciences & Electrical Engineering (Double Major), KAIST, Mar 2017 - Aug 2021

Experiences