Résumé |
In this talk I will review the relationship among robust physical
phenomena and our ability to describe them with approximate numerical
methods. I will use tensor networks as a paradigmatic example.
I will show that even though tensor network methods are designed to
work for slightly entangled states, they can be used as an approximate
tool in the scaling region of a quantum critical point. There the
constituents are strongly entangled.
Out of equilibrium, tensor networks algorithms typically fail due to
the large amount of entanglement that is generated among the
constituents. However, if we are able to identify the equivalent of a
scaling region, we can try to use tensor networks (or other
approximate algorithms) in order to characterize the universal part of
the dynamics.
I will discuss partial results in this direction. |