Explain anomalously large errors in odd radial-order modes

Investigate why the first two modes of each odd radial order exhibit larger errors than the other modes in the PO4NCPA reinforcement-learning correction of static non-common path aberrations using scalar and vector vortex coronagraphs.

Background

The paper evaluates PO4NCPA, a reinforcement-learning algorithm for focal-plane wavefront sensing and non-common path aberration correction, with scalar and vector vortex coronagraphs. In the static-aberration experiments, the authors observe that the first two Zernike modes of each odd radial order have larger residual errors in both coronagraphic configurations. They explicitly state that they do not have a clear explanation for this behavior and indicate that further analysis of the policy behavior is needed to identify its cause.

References

In both cases, we observe that the first two modes of each odd radial order display larger errors. So far, we do not have a clear explanation of this result and will need to further analyze the policy behavior for possible clues.

— Exploring reinforcement learning to enhance focal-plane wavefront control for vortex coronagraphs  (2609.01199 - Taskin et al., 1 Sep 2026) in Section 3.1, “Static NCPA”