additional_info.md

Additional figures

Instead of using uniformly distributed trajectories for training, here we use a special case of trajectories. i.e., starting point of the trajectories are at a fixed radial distance $R_0$ from the center.

Sample signal -- normalized, $s(t)/ \lvert s \rvert $

Weight profiles --- complete figure; Selected rows are shown in Figure 3 of the manuscript.

Effective measurement time --- complete figure; Selected lines are shown in Figure 3 of the manuscript.

Score shows the fraction of predictions that are accurate when using previously unseen data.