A Transformer-based learning-to-rank model for selected configuration interaction achieves chemical accuracy with substantially fewer determinants than prior classification or regression baselines across tested molecules.
Bulletin of Materials Science , volume=
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
physics.chem-ph 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Learning to Rank for Selected Configuration Interaction
A Transformer-based learning-to-rank model for selected configuration interaction achieves chemical accuracy with substantially fewer determinants than prior classification or regression baselines across tested molecules.