analiza aiqbench
This paper applies machine learning to materials science, specifically to finding adsorption sites and defect configurations in materials with little existing data. The PYTHIA framework combines a pretrained foundation model, frozen weights, a trainable transformer alignment layer, and Gaussian process uncertainty estimates, then uses the result as a surrogate for Bayesian optimization. For ML readers, the useful question is whether transfer learning from related chemical datasets generalizes to genuinely unexplored material families, which the abstract asserts but which will need to be checked against the benchmarks and baselines in the full paper.
PYTHIA, a transferable machine learning framework for simultaneous defect and adsorption site optimization
arXiv:2610.31539v1 Announce Type: new Abstract: Machine learning has permeated material optimization, providing high value. Nevertheless, its re- liance on a high number of samples limits its applicability to relatively explored material families. In this paper we extend the applicability of machine learning to novel and understudied materials through the PYTHIA (Probabilistic Yet Transferable Hierarchical Interaction Architecture) framework. Pythia utilizes two of the most powerful machine learning concepts, foundational models and transfer learning. We employ a complex foundational model trained on a large dataset of chemically similar molecules, on a target that is physically close to our prediction target. Then, we freeze the weights and perform subspace alignment with the use of a transformer layer and add uncertainty quantification with gaussian processes. These two layers can be trained on a limited dataset yet allow us to leverage the gained knowl- edge and the studied materials and complex architectures. The resulting model is used as a surrogate for a Bayesian Optimization scheme. To showcase the robustness of the framework, we applied it to the investigation of adsorption and defect optimization of two distinct systems. Not only were we able to correctly and efficiently identify the optimal adsorption sites of even complex a
Abstract:Machine learning has permeated material optimization, providing high value. Nevertheless, its re- liance on a high number of samples limits its applicability to relatively explored material families. In this paper we extend the applicability of machine learning to novel and understudied materials through the PYTHIA (Probabilistic Yet Transferable Hierarchical Interaction Architecture) framework. Pythia utilizes two of the most powerful machine learning concepts, foundational models and transfer learning. We employ a complex foundational model trained on a large dataset of chemically similar molecules, on a target that is physically close to our prediction target. Then, we freeze the weights and perform subspace alignment with the use of a transformer layer and add uncertainty quantification with gaussian processes. These two layers can be trained on a limited dataset yet allow us to leverage the gained knowl- edge and the studied materials and complex architectures. The resulting model is used as a surrogate for a Bayesian Optimization scheme. To showcase the robustness of the framework, we applied it to the investigation of adsorption and defect optimization of two distinct systems. Not only were we able to correctly and efficiently identify the optimal adsorption sites of even complex a
Machine Learning (cs.LG)
arXiv:2610.31539 [cs.LG]
(or arXiv:2610.31539v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.31539 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Text integral sub licența deschisă a publicației · 289 cuvinte
Subiecte conexe
Limitations on Joint Coherence Transfer in Quantum Thermodynamics
arXiv · 16 sept. 2026
Harnessing LLMs as Agents: What Does It Cost?
arXiv · 5 oct. 2026
Dongfeng and China Ordnance Group Plan 727 Million Yuan in Reciprocal Equity Swaps; Insider: Little Impact on Listed Companies' Core Businesses
36Kr · 20 oct. 2026
Researchers Cut Overhead in Quantum Error Assessment
Quantum Zeitgeist · 19 sept. 2026