Pushing the accuracy of on-top functionals with agent-driven supervised learning
Researchers introduced FunctionalAgent, an agentic workflow to systematically develop high-accuracy on-top functionals for multiconfiguration pair-density functional theory (MC-PDFT). This process involved automated dataset curation, quantum chemistry calculations, and iterative optimization. The resulting MC26 and COF26 functionals significantly outperformed previous methods by achieving superior performance across diverse chemical systems.