|Title||A neural-network approach to the many-body problem in open quantum system|
|Speaker||Alberto Biella (Université de Paris, LMPQ, CNRS)|
📅 Jul 18, 2019
|Location||SISSA, First Floor, Room 132|
Driven-dissipative systems have an intrinsic nonequilibrium nature and the properties of their stationary state cannot be determined via a free-energy analysis: collective phenomena and ordering are triggered by the simultaneous interplay between driving, dissipation and interactions. Their understanding is of fundamental importance to grasp the physics of a large variety of systems such as photonic quantum simulators (e.g. made of coupled QED cavities) and realistic quantum hardware where, in spite of the tremendous experimental progress, a perfect isolation from the environment is unavoidable and certainly plays a crucial role. In this context, the many-body problem has to account for the exponential growth of the underlying Hilbert space and the emergence of mixed-state dynamics due to the coupling to an external environment.
In this talk, I will review recent developments of this field with particular emphasis on numerical methods. In particular, I will discuss recent applications of neural network tools to simulate the behavior of an open many-body quantum system [1, 2] describing results and open challenges. Next, I will describe how these techniques allow one to study the phase-diagram of paradigmatic strongly-interacting dissipative spin [3, 4] and bosonic  systems. Particular attention will be devoted to the stabilization of exotic phases (without an equilibrium counterpart) and to the characterization of criticalities.
|🔗 Past seminars|
|🔗 Past journal club meetings|