CMT Group Seminar | October 28, 10:00
Tensor network calculation of classical frustrated systems
We discuss how the Tensor Network representation of classical statistical lattice models, with particular regards to the evaluation of the free energy, the entropy per site, and diverse correlation functions. We explain how the Corner Transfer Matrix (CTM) algorithm can be used for these sakes, and compare its results with those obtained via the Variational Uniform Matrix Product State (VUMPS) algorithm and the Tensor Renormalization Group (TRG) algorithm. In particular, we investigate whether tensor network methods are suitable to study frustrated systems, focussing here on the geometrically frustrated Ising antiferromagnetic model on the triangular lattice. We show that an accurate estimate of the non-trivial zero-point entropy can be obtained, and verify that the peculiar algebraic decay of the spin-spin correlation function with increasing distance can be faithfully reproduced. This gives us confidence to employ Tensor Networks to study classical spin liquids in the rest of this Master thesis.
https://uni-koeln.zoom.us/j/5601318943?pwd=UTFvNU5FeUZvU3E4dWdtRWo5bDh2dz09
Contact: not specified