Data from: Flatband localization and interactioninduced delocalization of photons
Data files
Sep 02, 2024 version files 1.36 MB

Figure2_Data.csv

Figure3_Data.csv

Figure4_Data.csv

README.md
Abstract
Lattices with dispersionless, or flat, energy bands have attracted significant interest in part due to the strong dependence of particle dynamics on interactions. Using superconducting circuits, we experimentally study the dynamics of one and two particles in a single plaquette of a lattice whose band structure consists entirely of flat bands. We first observe strictly localized dynamics of a single particle, the hallmark of allbandsflat physics. Upon initializing two particles on the same site, we see an interactionenabled delocalized walk across the plaquette. We further find localization in Fock space for two particles initialized on opposite sides of the plaquette. These results mark the first experimental observation of a quantum walk that becomes delocalized due to interactions and establishes a key building block in superconducting circuits for studying flatband dynamics with strong interactions.
README: Data from: Flatband localization and interactioninduced delocalization of photons
https://doi.org/10.5061/dryad.0cfxpnw80
The data included here was used to generate Figures 2, 3, and 4 in the main body of the text. Figure 2 characterizes single particle dynamics in the plaquette with zero and pi flux. In the pi flux device, we see localized dynamics where an initial excitation is localized to three of the four lattice sites. Figure 3 characterizes the dynamics of two particles initialized on the same site. Due to interactions, these particles hop as a pair and we find delocalized dynamics for the bound pair across the plaquette in both zero and pi flux devices. Figure 4 characterizes the dynamics of two particles initialized on opposite sides of the plaquette. Here, we find that when particles are initialized on the left and right sites (LR state), transfer to the TB state is prohibited in the pi flux plaquette.
Description of the data and file structure
There are three different CSV files for the three figures:
 Figure2_Data.csv
 Figure3_Data.csv
 Figure4_Data.csv
For the data in Fig. 2 and Fig. 3, the columns include "Time (us)" which provides the time axis from 0 to approximately 2.3 us in units of microseconds. Additionally, there are columns for each qubit, each device, experiment, and simulation. For example, the experimental data for the bottom qubit on the pi flux device is labeled "Qubit 3 PiF Exp". Qubit indeces 1, 2, 3, 4 correspond to labels "left", "top", "bottom", "right" in the paper. Simulation data uses the descriptor "Sim" rather than "Exp" and the zero flux device uses "0F" rather than "PiF".
Error bars are included for all experimental data. These error bars are 95% ClopperPearson confidence intervals and there are individual columns for the lower and upper intervals. Data was corrected for measurement errors using readout mitigation from qiskit python library.
There are a few data entries in the lower bound column of the datasets labeled as "n/a". This comes from how the ClopperPearson confidence intervals are generated: because the data is bounded below by zero, if the population is very low or zero, the lower bound will be invalid and therefore is registered as "n/a" or NaN in the data.
The data in Fig. 4 has the same columns and structure as for Fig. 2 and 3 with the exception of how qubits are designated. All measurements here were correlation measurements where the data from two qubits was simultaneously measured. The columns for the correlated data from the left (1) and right (4) qubit is labeled as Qubit 14. The column for experimental data in the zero flux device for the top (2) and bottom (3) qubits is "Qubit 23 0F Exp".
Methods
This data was collected by averaging 3000 single shot data points at each time evolution point. Data is bounded by 0 and 1 and was corrected for measurement errors using readout mitigation from the qiskit python library. The error bars for the data are 95% ClopperPearson confidence intervals.