UC Berkeley

The reliable determination of transition states (TSs) benefits from second-order information for robust convergence and validation, but the computational expense of Hessians prohibits their routine use in TS optimization. Here, we present a machine-learning-driven TS optimizer that directly predicts the leftmost Hessian eigenvector (LMHE), which is the critical mode that locally approximates the reaction coordinate encompassing the TS. We demonstrate that our LMHE optimizer recovers TS solutions at the same rate as full-Hessian optimizers and robustly from degraded initial guess geometries, thereby eliminating the excessively long wall times characteristic of full-Hessian approaches and reducing total gradient evaluations compared to standard quasi-Newton methods. We further improve accuracy and robustness using uncertainty quantification for identifying occasional LMHE prediction failures, where we then fall back to a full-Hessian update from the machine-learned potential at that optimization step, avoiding expensive active learning. Overall, our methodology and semiautomated workflow deliver second-order stability at first-order computational expense to provide a highly efficient engine for high-throughput reaction discovery.