Bidirectional Information Flow Quantum State Tomography

The exact reconstruction of many-body quantum systems is one the major challenges in modern physics, because it impractical to overcome exponential complexity problem brought by high-dimensional systems. Recently, machine learning techniques are well used promote information research and state tomography has also been developed neural network generative models. We propose a method, which based on bidirectional gated recurrent unit network, learn reconstruct both easy states hard this study. able use fewer measurement samples our method these obtain high fidelity.