Runtime Quantum Error Mitigation Using a Trained ML Model
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current quantum error mitigation techniques require additional execution time or access to more qubits for increased accuracy, leading to significant overhead and inefficiencies in quantum computing.
Innovation Solution
Training a machine learning model using noisy expectation values from parameterized circuits to perform quantum error mitigation at runtime, reducing overhead and execution time by learning the relationships between target and noisy expectation values.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If quantum error mitigation techniques (such as running additional mitigation circuits or modifying target circuits) are used to increase accuracy, then measurement precision is improved, but execution time increases
Solution Approach 1:
The patent applies preliminary action by training a machine learning model beforehand to learn the noise characteristics of the quantum device. This pre-trained model is then used at runtime to rapidly mitigate errors without requiring additional quantum circuit executions, thus improving measurement precision while avoiding the time penalty of traditional error mitigation techniques.
Solution Approach 2:
The patent uses copying by creating a classical machine learning model that replicates the noise behavior of the quantum device. Instead of running multiple quantum circuits to characterize and mitigate noise, a single classical model is trained to copy the noise characteristics, enabling fast error mitigation at runtime without additional quantum execution time.
2Measurement precision
If quantum error mitigation techniques are applied to reduce noise effects, then measurement precision is improved, but device complexity increases
Solution Approach 1:
The patent introduces an intermediary classical machine learning model that mediates between the noisy quantum device and the final results. This intermediary model handles the complexity of noise characterization and mitigation, allowing the quantum device itself to remain simple while achieving improved measurement precision through the classical post-processing step.
3Measurement precision
If traditional quantum error mitigation methods are used to achieve accurate expectation values, then measurement precision is improved, but overhead increases
Solution Approach 1:
The patent replaces the mechanical quantum circuit-based error mitigation approaches with a computational machine learning model. Instead of physically running additional quantum circuits to mitigate errors, the system uses a classical computational model to predict and correct noise effects, thereby reducing the overhead of mitigation circuits while maintaining measurement precision.
Data Source
AI summary
A method, system, and computer program product for runtime quantum error mitigation. Training data, which includes noisy expectation values and target expectation values (noiseless expectation values), is generated. A machine learning model is then trained using the training data to perform quantum error mitigation based on learning the relationships between target and noisy expectation values. That is, such a machine learning model is trained to generate target expectation values based on inputted noisy expectation values. Upon executing a quantum circuit on a quantum computer creating quantum results, quantum error mitigation is performed on the quantum results at runtime using the trained machine learning model. In this manner, there are significant savings in quantum execution time while improving the accuracy of the results in performing quantum error mitigation on quantum results at runtime without additional mitigation circuits.