Fig. 5: Comparison of Hamiltonian Rescaling Strategy with Richardson Extrapolation³⁴.

Although both the Hamiltonian rescaling strategy and Richardson extrapolation utilize the same experimental data, they differ fundamentally in their approach to error mitigation. a(1) In our Hamiltonian rescaling strategy, the signal from the experimental data is first processed to estimate the eigenvalue differences, followed by the application of the error mitigation procedure. a(2) In contrast, the Richardson extrapolation method mitigates the signal corresponding to the mean value of the measurement observable before estimating the eigenvalue differences using the mitigated signal. b Numerical comparison of these two strategies under the conditions described in Fig. 3. Given that the 〈O〉 vs. t signal behaves as a summation of damped oscillation modes, as discussed in “Results'', our Hamiltonian rescaling strategy demonstrates higher efficiency compared to the standard Richardson extrapolation, as shown in panel b. We note that the sample complexity of the Hamiltonian rescaling strategy is slightly higher than that of the Richardson extrapolation method (see Supplementary Information).