2026-10-11T12:09:36Z https://api.figshare.com/v2/oai

oai:figshare.com:article/33965070 2026-09-22T13:12:36Z category_1 category_4 category_915 category_15 category_21 category_272 category_873 category_931 category_133 category_135 portal_63 item_type_3 month_year_09_2026

ML26: New Ingredients for an Integral-Features Density Functional with Broad Chemical Accuracy Dayou Zhang (9037659) Yinan Shu (1475488) Benjamin G. Janesko (1441756) Donald G. Truhlar (1266384) Biophysics Biochemistry Physical Sciences not elsewhere classified Neuroscience Biotechnology Environmental Sciences not elsewhere classified Chemical Sciences not elsewhere classified Information Systems not elsewhere classified Plant Biology Computational Biology widely used mdb2019s layer neural network fixed hybrid percentage exhibits improved behavior improving upon ml25 future functional development learned density functionals features density functional global integral descriptors integral feature rather gmtkn55 data sets density functional tested functionals feature space earlier ml25 scalable framework projected rung nonlinear mapping local ingredients interaction benchmark enabling machine cs1 dynamic correlation ingredients conventional integration consistent accuracy challenging self based skala 89 kcal 5000 nodes 5 terms Integral-features density functional theory (IF-DFT) replaces the conventional integration over local ingredients with a nonlinear mapping from global integral descriptors, enabling machine-learned density functionals with broad transferability. Here we report ML26@MN15, a new IF functional that extends the earlier ML25@MN15 model by incorporating 12 additional correlation featuresincluding overlap-projected rung-3.5 terms, CS1 dynamic-correlation ingredients, and VV10 nonlocal correlationand by treating Hartree–Fock exchange as an integral feature rather than as a fixed hybrid percentage. ML26@MN15 employs 79 integral features evaluated on MN15 ingredients and a single-hidden-layer neural network with 5000 nodes to produce a size-extensive exchange–correlation energy. Trained on 185 databases comprising 7242 reference data, ML26@MN15 achieves a data-averaged energetic mean unsigned error of 0.89 kcal/mol, improving upon ML25@MN15 by 15% and outperforming a group of previous leading functionals across all eight chemical partitions examined. For the widely used MDB2019S, MGCDB84, and GMTKN55 data sets, ML26@MN15 yields the lowest averaged errors among all tested functionals, including the results for GMTKN55 by ML-based Skala-1.1 and DM21 density functional approximations. The new functional also maintains strong performance for systems containing heavy elements and exhibits improved behavior on a challenging self-interaction benchmark. These results show that expanding the integral-feature space with physically motivated nonlocal ingredients yields a density functional with unprecedented accuracy across diverse chemical benchmarks. The broad and consistent accuracy of ML26@MN15 highlights the promise of integral-features density functional approximations as a scalable framework for incorporating additional nonlocal physics into future functional development. 2026-09-22T00:00:00Z Dataset Dataset 10.1021/acs.jctc.6c01681.s001 https://figshare.com/articles/dataset/ML26_New_Ingredients_for_an_Integral-Features_Density_Functional_with_Broad_Chemical_Accuracy/33965070 CC BY-NC 4.0