Reproduce “MOFGPT: Generative Design of Metal-Organic Frameworks using Language Models” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Srivathsan Badrinarayanan, Rishikesh Magar, Akshay Antony, Radheesh Sharma Meda, Amir Barati Farimani
MOFGPT designs metal–organic frameworks for requested properties by generating MOFid strings and steering generations with property and validity feedback. A GPT-based generator samples candidates, a predictor scores them, and reinforcement learning adjusts the generator. For methane adsorption at 0.9 bar, the predicted mean rose from 0.818 to 1.834 mol/kg, below the 2.064 target; 53.42% were valid under the paper’s checks.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.