PreprintLicence not statedChemRxiv
Lost in Chemical Space: Local Open-Source Large Language Models Display Model-Dependent Adherence to Reference Data in Molecular Generation
- 1LNBio, CNPEM, R. Giuseppe Máximo Scolfaro, Campinas, São Paulo, Brazil, 13083-100
Related
- Analytic Energy Gradients for Unrestricted Mixed-Reference Spin-Flip Time-Dependent Density Functional Theory
- The Evolution of Electronic Structure in Conjugated Molecules Revealed by Mixed-Reference Spin-Flip Time-Dependent Density Functional Theory and Density-Based Analyses: II. Polyacenes
- How much of natural-product chemical space has a reference spectrum? An open audit of GNPS, MassBank and MoNA against COCONUT and LOTUS
- Matrix-dependent transformation governs the chemical identity and biological fate of a metal–organic framework
- A Machine Learning Approach to logBB Prediction Combining Molecular Docking and Quantum-Chemical Descriptors
BibTeX
@misc{bortot2026lost,
title = {{Lost in Chemical Space: Local Open-Source Large Language Models Display Model-Dependent Adherence to Reference Data in Molecular Generation}},
author = {Leandro Oliveira Bortot and Gabriel Filipe Totene and Lucas Abner Caixeta de Oliveira and Ariane Nunes-Alves},
howpublished = {ChemRxiv},
year = {2026},
month = nov,
doi = {10.26434/chemrxiv.15011763/v1},
url = {https://chemrxiv.org/doi/full/10.26434/chemrxiv.15011763/v1},
}
RIS
TY - UNPB
TI - Lost in Chemical Space: Local Open-Source Large Language Models Display Model-Dependent Adherence to Reference Data in Molecular Generation
AU - Leandro Oliveira Bortot
AU - Gabriel Filipe Totene
AU - Lucas Abner Caixeta de Oliveira
AU - Ariane Nunes-Alves
PY - 2026
DA - 2026/11/05
PB - ChemRxiv
DO - 10.26434/chemrxiv.15011763/v1
UR - https://chemrxiv.org/doi/full/10.26434/chemrxiv.15011763/v1
ER -