Assess selective rejection of loss configurations

Investigate whether discarding only selected loss configurations can reduce the loss-induced increase in the expected number of attempts for magic state cultivation without incurring unacceptable effects on acceptance or logical-error performance.

Background

The loss analysis assumes that all detectable loss events cause a shot to be discarded. Under independent loss channels, losses therefore increase the retry rate while leaving the logical error rate unchanged. The resulting multiplicative overhead can become substantial for larger cultivation circuits.

The authors observe that selectively discarding only some loss configurations might mitigate this overhead, but they do not determine which configurations should be rejected or quantify the resulting trade-offs. This is explicitly identified as future work.