Speaker
Description
Tensor renormalization group is expected to be a promising method to simulate lattice field theories at finite density since it does not suffer from the sign problem. We construct a Grassmann tensor network representing the partition function of 1+1D SU(2) lattice gauge theory coupled with staggered fermions. At finite couplings, a random sampling is applied to discretize the group integration. The initial bond dimension turns out to be $16K$ where $K$ is the number of SU(2) matrices sampled for each link variable. We introduce an efficient initial tensor compression scheme to reduce the size of initial tensors. Then, Grassmann bond-weighted tensor renormalization group approach is adopted to investigate a phase diagram in the $(m, \mu)$ plane with the quark mass $m$ and chemical potential $\mu$. The free energy density, number density, and diquark condensate at different gauge couplings are computed as a function of the chemical potential. We discuss the efficiency of random sampling method, our initial tensor compression scheme, and the future application toward the corresponding higher-dimensional models.
