# US 2025/0165825 A1 · Systems and methods for tensor-based variational quantum algorithm cost function landscape generation

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2025/0165825 A1
- Document ID: US20250165825A1
- Status: Published application
- Kind: A1
- Application number: 18/516,460
- Filed: 2023-11-21
- Published: 2025-05-22
- Earliest claimed: 2023-11-21
- Applicants: Jpmorgan Chase Bank, N.A.
- Inventors: Zichang HE; Ruslan SHAYDULIN; Marco PISTOIA
- Web page: https://patentgrep.com/patent/US20250165825A1

## CPC

- G06N 10/20: Models of quantum computing, e.g. quantum circuits or universal quantum computers
- G06N 10/60: Quantum algorithms, e.g. based on quantum optimisation, quantum Fourier or Hadamard transforms
- G06N 10/70: Quantum error correction, detection or prevention, e.g. surface codes or magic state distillation
- G06N: COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS

## Abstract

In some aspects, the techniques described herein relate to a method including: receiving a cost function, where the cost function is a function of parameters of a quantum circuit; discretizing the cost function to determine a number of dimensions and a number of elements in each dimension; formulating a landscape tensor, wherein the landscape tensor is formulated based on, and includes, the number of dimensions and the number of elements in each dimension; randomly sampling values of the parameters of the quantum circuit; executing the quantum circuit with the values of the parameters, wherein executing the quantum circuit generates values of the cost function; inserting the values of the cost function as values of corresponding elements in the number of elements included in the landscape tensor; and solving a low-rank tensor completion problem to estimate the values of empty elements in the landscape tensor.

## Claims

### Claim 1

A method comprising:
receiving, at a classical computing device, a cost function, where the cost function is a function of parameters of a quantum circuit;
discretizing the cost function to determine a number of dimensions and a number of elements in each dimension;
formulating a landscape tensor, wherein the landscape tensor is formulated based on, and includes, the number of dimensions and the number of elements in each dimension;
randomly sampling values of the parameters of the quantum circuit;
executing the quantum circuit with the values of the parameters, wherein executing the quantum circuit generates values of the cost function;
inserting the values of the cost function as values of corresponding elements in the number of elements included in the landscape tensor; and
solving a low-rank tensor completion problem to estimate the values of empty elements in the number of elements included in the landscape tensor.

### Claim 2 (depends on 1)

The method of claim 1, comprising:
storing the landscape tensor in a factorized form.

### Claim 3 (depends on 2)

The method of claim 2, wherein the factorized form includes only low-rank decomposed factors of the landscape tensor.

### Claim 4 (depends on 1)

The method of claim 1, comprising:
constructing a training data set from the values of the parameters of the quantum circuit.

### Claim 5 (depends on 4)

The method of claim 4, wherein the low-rank tensor completion problem minimizes an approximation error of the training data by solving low-rank factors.

### Claim 6 (depends on 5)

The method of claim 5, wherein the low-rank factors include {Ai}i=1d.

### Claim 7 (depends on 1)

The method of claim 1, wherein each element of the landscape tensor is represented by a vector of matrices.

### Claim 8

A system comprising at least one classical computing device including a processor and a memory, and at least one quantum computing device including a quantum processor, wherein the at least one classical computing device and the at least one quantum computing device are configured for operative communication with each other, and wherein the system is configured to:
receive, at the at least one classical computing device, a cost function, where the cost function is a function of parameters of a quantum circuit;
discretize the cost function to determine a number of dimensions and a number of elements in each dimension;
formulate a landscape tensor, wherein the landscape tensor is formulated based on, and includes, the number of dimensions and the number of elements in each dimension;
randomly sample values of the parameters of the quantum circuit;
execute the quantum circuit with the values of the parameters, wherein executing the quantum circuit generates values of the cost function;
insert the values of the cost function as values of corresponding elements in the number of elements included in the landscape tensor; and
solve a low-rank tensor completion problem to estimate the values of empty elements in the number of elements included in the landscape tensor.

### Claim 9 (depends on 8)

The system of claim 8, wherein the system is configured to:
store the landscape tensor in a factorized form.

### Claim 10 (depends on 9)

The system of claim 9, wherein the factorized form includes only low-rank decomposed factors of the landscape tensor.

### Claim 11 (depends on 8)

The system of claim 8, wherein the system is configured to:
construct a training data set from the values of the parameters of the quantum circuit.

### Claim 12 (depends on 11)

The system of claim 11, wherein the low-rank tensor completion problem minimizes an approximation error of the training data by solving low-rank factors.

### Claim 13 (depends on 12)

The system of claim 12, wherein the low-rank factors include {A i}d.

### Claim 14 (depends on 8)

The system of claim 8, wherein each element of the landscape tensor is represented by a vector of matrices.

### Claim 15

A non-transitory computer readable storage medium, including instructions stored thereon, which instructions, when read and executed by one of a classical computer processor or a quantum computer processor, cause the classical computer processor or the quantum computer processor to perform steps comprising:
receiving, at a classical computing device, a cost function, where the cost function is a function of parameters of a quantum circuit;
discretizing the cost function to determine a number of dimensions and a number of elements in each dimension;
formulating a landscape tensor, wherein the landscape tensor is formulated based on, and includes, the number of dimensions and the number of elements in each dimension;
randomly sampling values of the parameters of the quantum circuit;
executing the quantum circuit with the values of the parameters, wherein executing the quantum circuit generates values of the cost function;
inserting the values of the cost function as values of corresponding elements in the number of elements included in the landscape tensor; and
solving a low-rank tensor completion problem to estimate the values of empty elements in the number of elements included in the landscape tensor.

### Claim 16 (depends on 15)

The non-transitory computer readable storage medium of claim 15, comprising:
storing the landscape tensor in a factorized form.

### Claim 17 (depends on 16)

The non-transitory computer readable storage medium of claim 16, wherein the factorized form includes only low-rank decomposed factors of the landscape tensor.

### Claim 18 (depends on 15)

The non-transitory computer readable storage medium of claim 15, comprising:
constructing a training data set from the values of the parameters of the quantum circuit.

### Claim 19 (depends on 18)

The non-transitory computer readable storage medium of claim 18, wherein the low-rank tensor completion problem minimizes an approximation error of the training data by solving low-rank factors including {Ai}i=1d.

### Claim 20 (depends on 15)

The non-transitory computer readable storage medium of claim 15, wherein each element of the landscape tensor is represented by a vector of matrices.
