Figure 3
From: Constrained quantum optimization for extractive summarization on a trapped-ion quantum computer
The probability of sampling bitstrings with a given Hamming weight distance to the in-constraint subspace for 20 qubits. The Hamming weight distance is given by \(|\text{wt}(x)-M|\), where \(\text{wt}(x)\) is the Hamming weight of x and M is the constraint value. The shaded background represents the distribution of Hamming weight distance for random bitstrings sampled from a uniform distribution. Note that the overlap with the in-constraint subspace is lower on hardware due to the presence of noise.
