Accelerated Inorganic Materials Design with Generative AI Agents

Izumi Takahara, Teruyasu Mizoguchi, Bang Liu · 2025

Read Accelerated Inorganic Materials Design with Generative AI Agents by Izumi Takahara, Teruyasu Mizoguchi, Bang Liu (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-00741.txt.

Description

Izumi Takahara, Teruyasu Mizoguchi, and Bang Liu put a language model in charge of hunts for new inorganic stuff, with a diffusion model as its hands. Crystal design wants set traits, but new-form search can be slow, hard to read. Their agent, MatAgent, pairs a model that draws crystal forms with one that scores traits. Notes flow back and forth, and that loop steers the hunt to goals the user sets.

The agent also bears mind-like tools: short-term and long-term memory, the periodic table, and a broad base of facts on stuff. That frame mimics how a human pro thinks. Hunts land on the traits sought each time. Forms score high on sound build, on being rare, and on being new. The frame is a sound, broad path to new stuff, from batteries to catalysts.

Catalog record

CREATORIzumi Takahara, Teruyasu Mizoguchi, Bang Liu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE57 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2504-00741

Browse the full library · About the Tower of Babel