Choosing Threshold and Resetting Rate for Informed Stochastic Resetting

Determine criteria for selecting both the resetting threshold along a specified collective variable and the resetting rate in informed stochastic resetting for molecular dynamics and Metadynamics so as to yield efficient acceleration of mean first-passage times.

Background

The paper introduces informed stochastic resetting (ISR) as a way to accelerate molecular dynamics (MD) and Metadynamics (MetaD) by resetting trajectories only when predefined conditions tied to reaction progress are met (e.g., when a collective variable exceeds a threshold). ISR consistently outperforms standard stochastic resetting across model systems and even when using suboptimal collective variables.

While ISR demonstrates substantial speedups, practical deployment requires parameter choices—specifically, how to set the spatial threshold for resetting and how to choose an efficient resetting rate. The authors explicitly highlight this as an open question in the context of combining ISR with MetaD, motivating the development of predictive tools in the subsequent section to address it.