Guardado en:
Detalles Bibliográficos
Autores principales: Rubtsov, Ivan, Dudakov, Ivan, Kuratov, Yuri, Korolev, Vadim
Formato: Preprint
Publicado: 2025
Materias:
Acceso en línea:https://arxiv.org/abs/2511.03371
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866908630986522624
author Rubtsov, Ivan
Dudakov, Ivan
Kuratov, Yuri
Korolev, Vadim
author_facet Rubtsov, Ivan
Dudakov, Ivan
Kuratov, Yuri
Korolev, Vadim
contents Crystal graph neural networks are widely applicable in modeling experimentally synthesized compounds and hypothetical materials with unknown synthesizability. In contrast, structure-agnostic predictive algorithms allow exploring previously inaccessible domains of chemical space. Here we present a universal approach for enhancing composition-based materials property prediction by means of cross-modal knowledge transfer. Two formulations are proposed: implicit transfer involves pretraining chemical language models on multimodal embeddings, whereas explicit transfer suggests generating crystal structures and implementing structure-aware predictors. The proposed approaches were benchmarked on LLM4Mat-Bench and MatBench tasks, achieving state-of-the-art performance in 25 out of 32 cases. In addition, we demonstrated how another modeling aspect of chemical language models - interpretability - benefits from applying a game-theoretic approach, which is able to incorporate high-order feature interactions.
format Preprint
id arxiv_https___arxiv_org_abs_2511_03371
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Enhancing composition-based materials property prediction by cross-modal knowledge transfer
Rubtsov, Ivan
Dudakov, Ivan
Kuratov, Yuri
Korolev, Vadim
Materials Science
Computational Physics
Crystal graph neural networks are widely applicable in modeling experimentally synthesized compounds and hypothetical materials with unknown synthesizability. In contrast, structure-agnostic predictive algorithms allow exploring previously inaccessible domains of chemical space. Here we present a universal approach for enhancing composition-based materials property prediction by means of cross-modal knowledge transfer. Two formulations are proposed: implicit transfer involves pretraining chemical language models on multimodal embeddings, whereas explicit transfer suggests generating crystal structures and implementing structure-aware predictors. The proposed approaches were benchmarked on LLM4Mat-Bench and MatBench tasks, achieving state-of-the-art performance in 25 out of 32 cases. In addition, we demonstrated how another modeling aspect of chemical language models - interpretability - benefits from applying a game-theoretic approach, which is able to incorporate high-order feature interactions.
title Enhancing composition-based materials property prediction by cross-modal knowledge transfer
topic Materials Science
Computational Physics
url https://arxiv.org/abs/2511.03371