Features are disclosed for identifying and routing items for tagging using a latent feature model, such as a recurrent neural network language model (RNNLM). The model may be trained to identify latent features for catalog items such as movies, books, food items, beverages, and the like. Based on similarities in latent features, tags previous assigned to items may be applied to untagged items. Application may be manual or automatic. In either case, resources need to be balances to ensure efficient tagging of items. The included features help to identify and direct these limited tagging resources.

According to one embodiment, a calculating device includes a nonlinear oscillator. The nonlinear oscillator includes a circuit part including a first Josephson junction and a second Josephson junction, and a conductive member including a first terminal. An electrical signal is input to the first terminal. The electrical signal includes a first signal in a first operation. The first signal includes a first frequency component having a first frequency, and a second frequency component having a second frequency. The first frequency is 2 times an oscillation frequency of the nonlinear oscillator. An absolute value of a difference between the first frequency and the second frequency is not more than 0.3 times the first frequency.

A photonic processor uses light signals and a residue number system (RNS) to perform calculations. The processor sums two or more values by shifting the phase of a light signal with phase shifters and reading out the summed phase with a coherent detector. Because phase winds back every 2π radians, the photonic processor performs addition modulo 2π. A photonic processor may use the summation of phases to perform dot products and correct erroneous residues. A photonic processor may use the RNS in combination with a positional number system (PNS) to extend the numerical range of the photonic processor, which may be used to accelerate homomorphic encryption (HE)-based deep learning.

A photonic processor uses light signals and a residue number system (RNS) to perform calculations. The processor sums two or more values by shifting the phase of a light signal with phase shifters and reading out the summed phase with a coherent detector. Because phase winds back every 2π radians, the photonic processor performs addition modulo 2π. A photonic processor may use the summation of phases to perform dot products and correct erroneous residues. A photonic processor may use the RNS in combination with a positional number system (PNS) to extend the numerical range of the photonic processor, which may be used to accelerate homomorphic encryption (HE)-based deep learning.

In accordance with dissipative, photon-assisted quantum annealing described herein, a collection of qubits model a Boolean optimization problem, and the solution is determined by quantum annealing. However, rather than drive the qubits using a quasi-static field transverse to the computational direction, spins are allowed to evolve between computational states by multi-photon, inelastic collective scattering into a common waveguide coupled transversely to all of the qubits. Transitions between arbitrary states are enabled by the continuum of modes of the waveguide, while avoiding the exponential sensitivity to low-frequency decoherence near small gaps which is inherent in conventional QA. Moreover, because the transverse coupling to the waveguide averages to zero, the spin of each qubit experiences a net field purely in the computational direction, allowing continuous, quantum non-demolition measurement of the system.

In accordance with dissipative, photon-assisted quantum annealing described herein, a collection of qubits model a Boolean optimization problem, and the solution is determined by quantum annealing. However, rather than drive the qubits using a quasi-static field transverse to the computational direction, spins are allowed to evolve between computational states by multi-photon, inelastic collective scattering into a common waveguide coupled transversely to all of the qubits. Transitions between arbitrary states are enabled by the continuum of modes of the waveguide, while avoiding the exponential sensitivity to low-frequency decoherence near small gaps which is inherent in conventional QA. Moreover, because the transverse coupling to the waveguide averages to zero, the spin of each qubit experiences a net field purely in the computational direction, allowing continuous, quantum non-demolition measurement of the system.

A method for assembling a photonic computing system includes attaching a photonic source to a support structure, and attaching a photonic integrated circuit to the support structure. The photonic source includes a first laser die on a substrate configured to provide a first optical beam, and a second laser die on the substrate configured to provide a second optical beam. The photonic integrated circuit includes a first waveguide and a first coupler coupled to the first waveguide, and a second waveguide and a second coupler coupled to the second waveguide. The method includes attaching a plurality of beam-shaping optical elements to the support structure, the substrate, or the photonic integrated circuit, in which the attaching includes aligning a first beam-shaping optical element during attachment so that the first optical beam is coupled to the first coupler, and aligning a second beam-shaping optical element during attachment so that the second optical beam is coupled to the second coupler.

A method for assembling a photonic computing system includes attaching a photonic source to a support structure, and attaching a photonic integrated circuit to the support structure. The photonic source includes a first laser die on a substrate configured to provide a first optical beam, and a second laser die on the substrate configured to provide a second optical beam. The photonic integrated circuit includes a first waveguide and a first coupler coupled to the first waveguide, and a second waveguide and a second coupler coupled to the second waveguide. The method includes attaching a plurality of beam-shaping optical elements to the support structure, the substrate, or the photonic integrated circuit, in which the attaching includes aligning a first beam-shaping optical element during attachment so that the first optical beam is coupled to the first coupler, and aligning a second beam-shaping optical element during attachment so that the second optical beam is coupled to the second coupler.

A system for scalable, fault-tolerant photonic quantum computing includes multiple optical circuits, multiple photon number resolving detectors (PNRs), a multiplexer, and an integrated circuit (IC). During operation, the optical circuits generate output states via Gaussian Boson sampling (GBS), and the PNRs generate qubit clusters based on the output states. The multiplexer multiplexes the qubit clusters and replaces empty modes with squeezed vacuum states, to generate multiple hybrid resource states. The IC stitches together the hybrid resource states into a higher-dimensional cluster state that includes states for fault-tolerant quantum computation.

A system for scalable, fault-tolerant photonic quantum computing includes multiple optical circuits, multiple photon number resolving detectors (PNRs), a multiplexer, and an integrated circuit (IC). During operation, the optical circuits generate output states via Gaussian Boson sampling (GBS), and the PNRs generate qubit clusters based on the output states. The multiplexer multiplexes the qubit clusters and replaces empty modes with squeezed vacuum states, to generate multiple hybrid resource states. The IC stitches together the hybrid resource states into a higher-dimensional cluster state that includes states for fault-tolerant quantum computation.