Analog quantum simulations with Rydberg atoms in optical tweezers routinely address strongly correlated many-body problems due to the hardware-efficient implementation of the Hamiltonian. Yet, their generality is limited, and flexible Hamiltonian-design techniques are needed to widen the scope of these simulators. Here we report on the realization of spatially tunable interactions for XYZ models implemented by two-color near-resonant coupling to Rydberg pair states. Our results demonstrate the unique opportunities of Rydberg dressing for Hamiltonian design in analog quantum simulators.
Analog quantum simulations with Rydberg atoms in optical tweezers routinely address strongly correlated many-body problems due to the hardware-efficient implementation of the Hamiltonian. Yet, their generality is limited, and flexible Hamiltonian-design techniques are needed to widen the scope of these simulators. Here we report on the realization of spatially tunable interactions for XYZ models implemented by two-color near-resonant coupling to Rydberg pair states. Our results demonstrate the unique opportunities of Rydberg dressing for Hamiltonian design in analog quantum simulators.
Analog quantum simulations with Rydberg atoms in optical tweezers have emerged as a powerful tool for exploring strongly correlated many-body physics. However, the range of accessible phenomena remains limited by constraints in controlling many-body interactions in these systems. Here, we propose an alternative approach to realizing tunable interactions using Rydberg dressing on Aquila, a 256-atom quantum simulator based on optical tweezer arrays. By weakly admixing a Rydberg state via off-resonant laser coupling, we generate long-range interactions with temporal control. In combination with programmable atom geometries and site-selective control, this dressing protocol enables simulation of quantum spin models with state preparation and measurement in arbitrary basis. These capabilities enable access to exciting regimes of quantum simulation, including Quantum spin liquids, XYZ lattice models and broader application of quantum reservoir computing. Our approach establish Rydberg dressing as a practical and scalable tool for expanding the class of many-body systems accessible to analog quantum simulation and provides a critical stepping stone toward quantum machine learning and hybrid analog-digital quantum simulation with neutral atoms.
We use the resonant dipole-dipole interaction between Rydberg atoms and a periodic external microwave field to engineer XXZ spin Hamiltonians with tunable anisotropies.The atoms are placed in one-dimensional (1D) and two-dimensional (2D) arrays of optical tweezers.As illustrations, we apply this engineering to two iconic situations in spin physics: the Heisenberg model in square arrays and spin transport in 1D.We first benchmark the Hamiltonian engineering for two atoms and then demonstrate the freezing of the magnetization on an initially magnetized 2D array.Finally, we explore the dynamics of 1D domain-wall systems with both periodic and open boundary conditions.We systematically compare our data with numerical simulations and assess the residual limitations of the technique as well as routes for improvement.The geometrical versatility of the platform, combined with the flexibility of the simulated Hamiltonians, opens up exciting prospects in the fields of quantum simulation, quantum information processing, and quantum sensing.
Quantum simulators have the potential to solve quantum many-body problems that are beyond the reach of classical computers, especially when they feature long-range entanglement. To fulfill their prospects, quantum simulators must be fully controllable, allowing for precise tuning of the microscopic physical parameters that define their implementation. We consider Rydberg-atom arrays, a promising platform for quantum simulations. Experimental control of such arrays is limited by the imprecision on the optical-tweezer positions when assembling the array, hence introducing uncertainties in the simulated Hamiltonian. In this work, we introduce a scalable approach to Hamiltonian learning using graph neural networks (GNNs). We employ the density-matrix renormalization group to generate ground-state snapshots of the transverse-field Ising model realized by the array, for many realizations of the Hamiltonian parameters. Correlation functions reconstructed from these snapshots serve as input data to carry out the training. We demonstrate that our GNN model has a remarkable capacity to extrapolate beyond its training domain, regarding both the size and the shape of the system, yielding an accurate determination of the Hamiltonian parameters with a minimal set of measurements. We prove a theorem establishing a bijective correspondence between the correlation functions and the interaction parameters in the Hamiltonian, which provides a theoretical foundation for our learning algorithm. Our work could open the road to feedback control of the positions of the optical tweezers, hence providing a decisive improvement of analog quantum simulators.
Optical tweezer arrays of neutral atoms have emerged as a versatile platform for analog quantum simulation and digital quantum computation, combining scalability, reconfigurable geometry, and flexible microscopic control. In the first part of this work, we develop a 133Cs Rydberg simulator and demonstrate coherent Rydberg excitation in a rearrangeable array of more than 100 atoms. We then use this platform to probe critical many-body physics, extracting the scaling dimensions of the Ising universality class from an adiabatically prepared critical state. Dual-species arrays extend these capabilities by introducing spectroscopic selectivity and independently tunable interspecies interactions. Motivated by these opportunities, we develop a dual-species Rydberg array of 23Na and 133Cs atoms. The large separation between the optical transitions of the two species allows qubits encoded in one species to be cooled, trapped, imaged, and excited while the other species remains largely undisturbed. At the same time, strong van der Waals interactions provide a programmable coupling between the two arrays. We use this platform to demonstrate a proof-of-principle multi-qubit stabilizer readout based on an interaction-compensated protocol. In this setting, spectroscopic separation becomes an architectural advantage: it enables in-situ ancilla control and readout without requiring atom transport between separate interaction and measurement zones. Consequently, a multi-qubit entangling operation that would otherwise be decomposed into a sequence of two-qubit gates can be implemented as a native, effectively constant-depth primitive. These results point toward faster and more scalable syndrome extraction in neutral-atom quantum error correction, and establish dual-species Rydberg arrays as a promising architecture connecting programmable quantum simulation with fault-tolerant quantum computation.
Neutral atoms trapped and manipulated by laser light provide experimentally well accessible quantum systems allowing for a high degree of control over external and internal degrees of freedom. Arrays of dipole traps in which the atoms are confined individually in a configurable geometry constitute a versatile platform for quantum simulation and information applications. By exciting these atoms into Rydberg states, interactions of variable strength and range can be introduced into the system, allowing for the implementation of entangling gate operations or spin Hamiltonians. A crucial requirement for these schemes to function in a reliable way is the ability to create defect-free arrays of single atoms. This is generally a challenge in these types of systems, as common atom loading schemes are limited to roughly 50% probability of filling each site. In this work, a technique for the rearrangement of atoms within a scalable architecture based on micro-optical lens arrays was developed and implemented, resulting in the creation of uniformly filled regions containing more than 100 atoms, which represent the largest defect-free structures realized so far in systems of this kind. This was accomplished by filling the empty traps in a pre-defined pattern with an atom one by one using an optical tweezer. Thus, structures with up to 5 x 5 atoms could be rendered defect-free in more than 99% of attempts. Although the success rate drops below unity for larger clusters, a value of 3.1% for a 100-atom structure is still viable for experiments working with post-selection methods. The filling fraction of even the largest examined structures was observed to be higher than 88%, surpassing common loading schemes by a significant margin. The measurements presented in this thesis build on a region of the array containing 361 sites, being limited by available laser power. In contrast, the addressable range of the optical tweezer includes more than 1500 sites and microlens arrays with up to a million lenses are commercially available. By implementing coherent Rydberg excitation of this assembled atom array, significant progress toward a universal quantum computer or flexible quantum simulator has been made. Using a two-photon excitation scheme, coherent dynamics between the ground and Rydberg state could be observed simultaneously in a 5 x 5 region of the array, with two-photon Rabi frequencies on the order of Omega = 2Pi x 500kHz measured for a Rydberg laser beam waist of w_0,B = 18.7(10) µm. The choice of an appropriate Rydberg state and interatomic spacing led to the presence of strong nearest-neighbor interactions and allowed for the demonstration of the Rydberg blockade effect by observing a collective enhancement of the Rabi frequency consistent with the expected scaling ~ sqrt(N) as well as the suppression of multiple excitations. This mechanism represents the fundamental constituent of a two-qubit gate operation. The architecture introduced in this work offers a scalability unique among quantum simulation platforms and the presented results underpin its potential to propel the atom-optical approach for quantum information processing beyond the threshold of quantum supremacy. Different approaches for scaling up the system have been explored, indicating that defect-free structures of more than 1000 atoms are within range with feasible experimental improvements. Through a detailed analysis of the factors limiting the coherence of the observed dynamics, strategies for future experimental improvements have been developed. Among these, increasing the coupling strength to the Rydberg state into the megahertz regime by increasing laser power and implementing single-site addressing represents the most straight-forward and promising approach.