How is sensitivity analysis conducted in pybamm

[YoungAgile]

We can do sensitivity analusis in this way referring to https://github.com/pybamm-team/workshop-notebooks/blob/main/2021-09-workshop/Session%203/sensitivities_and_data_fitting.ipynb

solution = solver.solve(
    model, t_eval,
    inputs={"Current function [A]": 0.15652},
    calculate_sensitivities=True
)

I want to know How is sensitivity analysis conducted in pybamm. What method was used.

[YoungAgile]

And why does sensitivity change over time