Computing More with Less: Random Tensor Networks and New Algorithms for Time Evolution
Jacopo De Nardis (Université Paris-Cergy)
I will present recent analytical and numerical results on quantum complexity and anticoncentration in many-body wave functions, using random tensor-network ensembles as a powerful, tractable testbed for non-equilibrium dynamics. I will then introduce new algorithms for the classical simulation of quantum time evolution: (i) hybrid tensor-network methods enhanced by Clifford unitaries, and (ii) Monte Carlo sampling of spatially contracted networks. Together, these approaches extend the reach of tensor network simulations, showing how less resources suffice to capture key quantum features.