Start this stage from the model’s declared initial conditions rather than the previous stage’s endpoint. The update rules’ tuned targets always cross the stage boundary; what resets is the dynamical state — state variables, delay and coupling history, observation window buffers and monitors — and the noise stream, which restarts from the run’s seed so each stage samples the same realisation. False, the default, makes the schedule one continuous trajectory; true makes each stage an independent simulation warm-started from the parameters learned so far.
name: reset_statedescription: Start this stage from the model's declared initial conditions rather than the previous stage's endpoint. The update rules' tuned targets always cross the stage boundary; what resets is the dynamical state — state variables, delay and coupling history, observation window buffers and monitors — and the noise stream, which restarts from the run's seed so each stage samples the same realisation. False, the default, makes the schedule one continuous trajectory; true makes each stage an independent simulation warm-started from the parameters learned so far.from_schema: https://w3id.org/tvborank:1000ifabsent:'False'owner: AlgorithmStagedomain_of:- AlgorithmStagerange: boolean