Fig. 6
TTS for QAC and the C strategy, sorted by added noise. We show the TTS to find at least one ground state of the median instance of class \((L,\eta )\). In a, we show the results from the classical repetition strategy. For large instances \(L\ge 10\) and high noise \(\eta =0.15\), the C strategy fails to find any ground state for the median instance in our data set. In b, we show the results for QAC, which is always able to find the ground state of the median instance in our data set. The scaling in b is milder compared to a. For all \(\eta \ >\ 0\), a missing data point indicates that the ground state was never found at that \(L\) value, for example., for all \(L\ge 14\) in a.
