Quantum State Fidelity Lower Bound via SVD Truncation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current approximated methods for quantum systems, such as tensor network-based models, lack a method to determine the closeness of approximated final quantum states to exact final quantum states, limiting their usefulness in understanding complex systems.
Innovation Solution
A method is introduced to calculate a lower bound of the fidelity between an approximated final quantum state and an exact final quantum state by iteratively applying quantum gates, factorizing the quantum state using singular value decomposition, and truncating the bond dimensions to maintain a predetermined threshold, allowing for the determination of truncation fidelities and a product-based lower bound of fidelity.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If tensor network-based approximated methods are used to study quantum systems, then the complexity is reduced and larger systems can be handled, but there is no method to determine if the approximated final quantum state is close to the exact final quantum state
Solution Approach 1:
The patent introduces an intermediary metric called 'truncation fidelity' that mediates between the approximated quantum state and the exact quantum state. This intermediary allows indirect measurement of the quality of approximation by tracking the fidelity loss at each truncation step, thereby enabling fidelity assessment without directly computing the exact final quantum state.
Solution Approach 2:
The patent applies preliminary action by calculating and tracking truncation fidelity at each intermediate step of the tensor network evolution, rather than attempting to assess fidelity only at the final state. This preliminary tracking of fidelity metrics throughout the computation allows for continuous quality assessment and enables early detection when approximations become unacceptable.
2Productivity
If bond dimension truncation is applied to maintain a maximum bond dimension, then the computational tractability is maintained, but the accuracy of the quantum state representation is reduced
Solution Approach 1:
The patent implements feedback by using the calculated truncation fidelity as a feedback signal to guide the truncation process. The truncation fidelity information feeds back into the decision-making process for bond dimension management, allowing dynamic adjustment of truncation thresholds and enabling operators to assess whether the current level of approximation is acceptable for their specific application.
Solution Approach 2:
The patent applies parameter changes by dynamically adjusting the bond dimension threshold and truncation criteria based on the calculated truncation fidelity. Instead of using a fixed bond dimension limit, the method allows the parameters controlling approximation quality to be modified based on the observed fidelity loss, thereby optimizing the balance between computational efficiency and accuracy.
3Ease of operation
If no fidelity determination method is available, then approximated methods can be applied without verification, but the physical meaningfulness of the results cannot be confirmed
Solution Approach 1:
The patent enables self-service by providing a built-in fidelity assessment mechanism that operates automatically within the tensor network framework. The truncation fidelity calculation is integrated into the existing computational workflow, allowing the method to self-assess the quality of its own approximations without requiring external verification tools or additional computational overhead beyond the natural evolution process.
Data Source
AI summary
Method for determining a lower bound of a fidelity of an approximated final state, comprising:receiving an initial state in a matrix product representation;receiving a quantum circuit comprising gates;iterating over the gates:applying a current gate to the initial state;if the current gate is a two-qubit gate, factorizing a portion of the updated state by SVD into a product of a unitary matrix, a diagonal matrix, and a unitary matrix;if a bond dimension of the diagonal matrix exceeds a threshold: truncating the diagonal matrix such that the bond dimension does not exceed said threshold, and determining a truncation fidelity;in a next iteration, using the updated state as the initial state;determining a lower bound of the fidelity of the approximated final state as a product of the truncation fidelities.