We review the basic ideas of the tensor renormalization group method and show how they can be applied for lattice field theory models involving relativistic fermions and Grassmann variables in arbitrary dimensions. We discuss recent progress for entanglement filtering, loop optimization, bond-weighting techniques and matrix product decompositions for Grassmann tensor networks. The new methods are tested with two-dimensional Wilson-Majorana fermions and multi-flavor Gross-Neveu models. We show that the methods can also be applied to the fermionic Hubbard model in 1+1 and 2+1 dimensions.
Abstract We derive a general formula of the tensor network representation for d-dimensional lattice fermions with ultra-local interactions, including Wilson fermions, staggered fermions, and domain-wall fermions. The Grassmann tensor is concretely defined with auxiliary Grassmann variables that play a role in bond degrees of freedom. Compared to previous works, our formula does not refer to the details of lattice fermions and is derived by using the singular value decomposition for the given Dirac matrix without any ad-hoc treatment for each fermion. We numerically test our formula for free Wilson and staggered fermions and find that it properly works for them. We also find that Wilson fermions show better performance than staggered fermions in the tensor renormalization group approach, unlike the Monte Carlo method.
We present a tensor-network approach for two-dimensional strong-coupling QCD with staggered quarks at nonzero chemical potential. After integrating out the gauge fields at infinite coupling, the partition function can be written as a full contraction of a tensor network consisting of coupled local numeric and Grassmann tensors. To evaluate the partition function and to compute observables, we develop a Grassmann higher-order tensor renormalization group method, specifically tailored for this model. During the coarsening procedure, the blocking of adjacent Grassmann tensors is performed analytically, and the total number of Grassmann variables in the tensor network is reduced by a factor of two at each coarsening step. The coarse-site numeric tensors are truncated using higher-order singular value decompositions. The method is validated by comparing the partition function, the chiral condensate and the baryon density computed with the tensor method with exact analytical results on small lattices up to volumes of 4×4. For larger volumes, we present first tensor results for the chiral condensate as a function of the mass and volume, and observe that the chiral symmetry is not broken dynamically in two dimensions. We also present tensor results for the number density as a function of the chemical potential, which hint at a first-order phase transition.
The phase diagram of QCD is a central research topic in modern particle physics. Its calculation, however, is notoriously difficult especially when not only nonzero temperatures
T but also non-vanishing baryon densities are considered – as is relevant for heavy‑ion collisions or inside large neutron stars. At the theoretical level, nonzero baryon densities are achieved by introducing a nonzero chemical potential
μ, which creates a particle-antiparticle asymmetry. It turns out that for nonzero chemical potential, the fermion determinant is no longer real but complex, and therefore standard approaches based on importance sampling break down. In the literature, this is commonly referred to as the sign problem. In order to circumvent this sign problem, a variety of methods has been proposed. Examples include a Taylor expansion in
μ, reweighting, complex Langevin, thimbles or path optimization. Nevertheless, none of the proposed remedies can successfully reach regimes in which
μ/T>1, and therefore even today the full QCD phase diagram in the
μ−T plane cannot be calculated. There is a promising further approach, called tensor-network approach, that differs crucially from all of the aforementioned approaches. In this approach, the partition function is written as a contraction of a large number of local tensors. Of special relevance is the higher order tensor renormalization group (HOTRG) approach, which can be applied for arbitrary number of dimensions, and in particular its generalization, the Grassmann HOTRG (GHOTRG) approach, which is applicable to theories including fermions. Recently, GHOTRG was tailored to strong-coupling QCD using one flavor of staggered quarks in two dimensions. It was observed that, for a non‑zero chemical potential, no numerical instabilities arise. In this thesis, we go beyond both the strong-coupling limit and the single-flavor case. We express the partition function of lattice QCD with staggered fermions in the strong-coupling expansion in terms of a tensor network. This is done for an arbitrary number of dimensions, flavors, and orders of the strong-coupling expansion. Further, the GHOTRG approach is extended such that every coefficient
Zn of the partition function
Z=∑nZnβn can be calculated separately. Finally, numerical results in two dimensions are studied.