Dynamic LOCC Circuits for Automated Entanglement Manipulation

Tue, 09 Sep 2025 00:00:00 +0000

LCQNN: Linear Combination of Quantum Neural Networks

Thu, 03 Jul 2025 00:00:00 +0000

Quantum Entanglement Allocation through a Central Hub

Thu, 12 Sep 2024 00:00:00 +0000

Optimal unilocal virtual quantum broadcasting

Wed, 24 Jul 2024 00:00:00 +0000

Mitigating Barren Plateaus of Variational Quantum Eigensolvers

Fri, 29 Mar 2024 00:00:00 +0000

Paper accepted by IEEE Transactions on Quantum Engineering!

Thu, 14 Mar 2024 00:00:00 +0000

Paper on the barren plateaus of VQE is accepted by IEEE Transactions on Quantum Engineering 🎉

Our paper Mitigating barren plateaus of variational quantum eigensolvers by Xia Liu, Geng Liu, Haokai Zhang, Jiaxin Huang and Xin Wang was accepted by IEEE Transactions on Quantum Engineering!

For IEEE Transactions on Quantum Engineering: IEEE Transactions on Quantum Engineering (TQE) focuses on the interdisciplinary field of quantum engineering. It serves as a platform for researchers and practitioners to publish and exchange cutting-edge research on various aspects of quantum information science and technology. TQE aims to advance the understanding and development of quantum technologies, including quantum computing, quantum communication, quantum cryptography, and quantum sensing.

For the paper Mitigating barren plateaus of variational quantum eigensolvers: Variational quantum eigensolvers (VQE) is acknowledged as a promising class of variational quantum algorithms to solve classically intractable problems in physical and chemical systems. However, the barren plateau phenomenon seriously hinders the performance of the algorithm, that is, the gradient of the cost function vanishes exponentially with the system size so that the training process will be hampered for a moderately large system. In this work, we propose the state efficient ansatz (SEA) based on quantum information theory for accurate ground state preparation with improved trainability. We prove that barren plateaus can be efficiently mitigated by the SEA and the trainability can be further improved at most quadratically by flexibly adjusting the entangling capability of the SEA. From the perspective of frame potential, we demonstrate that SEA maintains high trainability without sacrificing too much expressibility, allowing for a trade-off between trainability and expressibility. Numerical simulations show that SEA obtains significant improvements in the magnitude of cost gradient and the convergence speed. Our results broaden a new perspective to address BP by balancing the trade-off between expressibility and trainability.

Xia Liu | Wang Research Group

Dynamic LOCC Circuits for Automated Entanglement Manipulation

LCQNN: Linear Combination of Quantum Neural Networks

Quantum Entanglement Allocation through a Central Hub

Optimal unilocal virtual quantum broadcasting

Mitigating Barren Plateaus of Variational Quantum Eigensolvers

Paper accepted by IEEE Transactions on Quantum Engineering!