Jihyeon Hur

I am a PhD student at KAIST, advised by Prof. Noseong Park at BigDyL. I received my MS from Seoul National University, advised by Prof. Gahgene Gweon, and my BS in Computer Science and Engineering from Jeonbuk National University.

I am interested in Scientific Machine Learning and LLMs for mathematics. My research focuses on neural operators for solving PDEs and training LLMs to tackle research-level mathematical problems.

Email  /  CV  /  Scholar  /  Github

Publications

An adaptive memory gating mechanism for neural operators that dynamically controls memory usage for PDE solving.

Addresses the distribution gap between synthetic and real dialogue data for effective data augmentation.

Human–LLM co-evaluation framework for assessing student collaboration quality in data-centric programming courses.

A pitch contour model with Transformer cross-attention that captures prosodic features for improved speech emotion recognition.

Self-supervised point cloud upsampling leveraging pretrained denoising models, without requiring paired training data.

Point cloud upsampling method using deep self-sampling guided by point saliency to focus on geometrically important regions.

A comprehensive survey of recent deep learning approaches for point cloud upsampling.

Transformer-based model for segmenting and monitoring fruit growth in peach tree imagery.

Benchmarking study of deep learning-based object detectors for automated apple harvesting systems.

Template from Jon Barron.