Benchmarking Large Language Models for Polymer Property Predictions
Macromolecular Rapid Communications, EarlyView.Large language models (LLMs) are fine‐tuned on polymer thermal property datasets to directly predict glass transition, melting, and decomposition temperatures from SMILES inputs. Compared to state‐of‐the‐art models such as Polymer Genome, polyGNN, and polyBERT, LLMs achieve competitive yet lower accuracy.
Sonakshi Gupta +3 more
wiley +1 more source
Glucose/potassium ratio and a novel combined model for the differential diagnosis of testicular torsion: a retrospective study. [PDF]
Basic Clin AndrolAktaş Y, Tunçekin A.
europepmc +1 more source
Semantic Embeddings of Chemical Elements for Enhanced Materials Inference and Discovery
Materials Genome Engineering Advances, EarlyView.ElementBERT extracts semantic embeddings of chemical elements from 1.29 million alloy‐related abstracts, providing robust descriptors that improve prediction accuracy by up to 23% across titanium, high‐entropy, and shape memory alloys, with demonstrated generalization on alloy compositions reported in 2025.
Yunze Jia +7 more
wiley +1 more source
Machine Learning-Based Prediction of Textural Properties and Nonlinear Regulatory Pattern Analysis of 3D-Printed Dough Containing Konjac Glucomannan. [PDF]
FoodsLeng W, Sun Y, Xie J, Pang J.
europepmc +1 more source
, 2000 Mobile telephony has been for many years the most popular application supported by mobile systems such as the Global System for Mobile communications (GSM). Recently, the use of mobile data applications such as the GSM Short Message Service (SMS) has gained popularity.
openaire +1 more source
Materials Genome Engineering Advances, EarlyView.An uncertainty‐aware FEM–GPR active learning framework efficiently explores the design space of 3D braided W–Cu composites, identifying architectures with higher yield strength and preserved electrical conductivity, whereas SHAP interpretation highlights yarn spacing in the braid plane as the key structural factor controlling composite performance ...
Lai Zhang +6 more
wiley +1 more source
In-vivo dosimetry using electronic portal imaging in online-adaptive radiotherapy for rectal cancer. [PDF]
Phys Imaging Radiat OncolRazinskas G +7 more
europepmc +1 more source
Sundman‐Like Transformations and the NRT Nonlinear Schrödinger Equation
Mathematical Methods in the Applied Sciences, EarlyView.ABSTRACT We present a new generalization of the well‐known power‐type Sundman transformation, involving not only powers of the function but also of its derivative, along with its inverse. Our aim is to explore the use of such transformations in the derivation of solutions of ordinary differential equations and in the study of their properties.
P. R. Gordoa +3 more
wiley +1 more source
RBF-assisted surrogate modeling and machine learning for mechanical property prediction in friction stir additive manufacturing: Application to dissimilar AA6061/AA7075 aluminum alloys. [PDF]
Sci RepVenkatachalam K +4 more
europepmc +1 more source
International Journal of Mechanical System Dynamics, EarlyView.ABSTRACT Understanding the dynamic behavior of structural components is crucial for optimizing performance and ensuring structural integrity. This study presents a new method that combines a systematic experimental investigation of four distinct hole geometries (circular, square, compact rectangular, and long rectangular) with varying hole counts, all ...
Amir Hossein Rabiee +3 more
wiley +1 more source