Hyunsoo Park

Materials.AI Researcher at Imperial College London

I am a Full Stacker of "Computational Materials Design"

Bridging Materials ScienceComputational ChemistryDeep Learning

Google Scholar LinkedIn GitHub

Journey

Imperial College London

2023-Present

Postdoc (2023.09 - Present)

Imperial College London

Generative AIMLFFAgentic AI

London, United Kingdom

Main Projects

Chemeleon2: GRPO for generative modeling for materials (arXiv, 2025)
DAC-SIM: A molecular simulation package integrating MLFFs in MOFs for DAC (Matter, 2025)
Chemeleon: A text-guided denoising diffusion model for materials generation (Nature Communications, 2025)

Perspectives

Closing the synthesis gap in computational materials design (arXiv, 2025)
Reflections on Europe academia as a postdoc (Matter, 2025)
Has generative artificial intelligence solved inverse materials design? (Matter, 2024)

Awards

Won 1st place in 2025 KRICT ChemDX Hackathon (MatSX)
Won 1st place in Global LLM Hackathon for Applications in Materials Science & Chemistry (Awards)

Advisor: Aron Walsh

Korea Advanced Institute of Science and Technology (KAIST)

2020-2023

PhD (2020.03 - 2023.08)

Korea Advanced Institute of Science and Technology (KAIST)

MOFTransformerTransfer LearningMOFs

Daejeon, South Korea

Main Project

MOFTransformer: A multi-modal transformer for universal transfer learning in MOFs (Nature Machine Intelligence, 2023)

Advisor: Jihan Kim

EPFL

2022

Visiting Researcher (2022.05 - 2022.10)

EPFL

Reinforcement LearningInverse DesignMOFs

Sion, Switzerland

Main Project

MOFReinforce: RL framework for inverse design of MOFs for CO2 DAC application (Digital Discovery, 2024)

Advisor: Berend Smit

Featured

Chemeleon2

Chemeleon2

arXiv preprint (2025)

DAC-SIM

DAC-SIM

Matter (2025)

Chemeleon

Chemeleon

Nature Communications (2025)

LLM Hackathon

LLM Hackathon

Global LLM Hackathon (2025)

MOFTransformer

MOFTransformer

Nature Machine Intelligence (2023)

Tutorial

MLFF Tutorial

Jupyter Notebook

MACE fine-tuning for Metal-Organic Frameworks

Open in GitHub →

Chemeleon Tutorial

Crystal discovery pipeline: SMACT, Chemeleon-DNG, MACE, Stability Analysis

Open in Colab →

Chemeleon2 Documentation

VAE, LDM, and RL modules for crystal structure generation with GRPO

Blog

TBD

To Be Soon

Template blog post for future updates.

Paper

Selected Papers

1. Hyunsoo P.^†, Aron W.^† Guiding Generative Models to Uncover Diverse and Novel Crystals via Reinforcement Learning. arXiv preprint (2025).

2. Yunsung L.*, Hyunsoo P.*, Aron W. and Jihan K. Achieving high-throughput screening of CO₂ Direct Air Capture materials with a transferable machine learning force field. Matter. 8, 102203 (2025).

3. Hyunsoo P.^†, Anthony O. and Aron W.^† Exploration of crystal chemical space using text-guided generative artificial intelligence. Nature Communications. 16, 4379 (2025).

4. Yeonghun K.*, Hyunsoo P.*, Berend S., and Jihan K. A Multi-modal Pre-training Transformer for Universal Transfer Learning in Metal-Organic Frameworks. Nature Machine Intelligence. 5, 309-318 (2023).

5. Hyunsoo P., Zhenzhu L. and Aron W. Has generative artificial intelligence solved inverse materials design?. Matter. 7, 2355-2367 (2024).

6. Hyunsoo P., Kinga O M., Panyalak D., Ryan N., Aron W. Closing the synthesis gap in computational materials design. ChemRxiv preprint (2025).

Get in Touch

Feel free to reach out for collaborations, questions, or just to say hi.

phs68660888@gmail.com