An LLM pipeline extracts thermodynamic data from literature and the resulting dataset trains a CatBoost model that predicts mineral formation enthalpy with a mean absolute error of 46.5 kJ/mol.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cond-mat.mtrl-sci 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Thermodynamic Prediction Enabled by Automatic Dataset Building and Machine Learning
An LLM pipeline extracts thermodynamic data from literature and the resulting dataset trains a CatBoost model that predicts mineral formation enthalpy with a mean absolute error of 46.5 kJ/mol.