We investigate a class of four-fermion theories which includes well-known models like the Gross-Neveu model and the Thirring model. In three spacetime dimensions, they are used to model interesting solid state systems like high temperature superconductors and graphene. Additionally, they serve as toy models to study chiral symmetry breaking (CSB).

For any number of fermion flavours the Gross-Neveu model has a broken and a symmetric phase, while the existence of a broken phase in the Thirring model depends on the number of flavours. The critical number of fermion flavours beyond which there exists no CSB is still subject of ongoing discussions. Using SLAC fermions we simulate the Thirring model with exact chiral symmetry. To obtain a chiral condensate one can introduce a symmetry-breaking mass term and carefully study the limits of infinite lattice and zero-mass. So far, we did not see CSB within this approach for the Thirring model with 2 or more (reducible) flavours.

The talk presents alternative approaches to investigate these findings. We employ certain Fierz identities to map the Thirring model into equivalent four-fermion models, for which the chiral condensate does not seem to vanish. In the new formulations based on reshuffled degrees of freedom we find a sign problem (which is not present in the original formulation). For this reason we developed an algorithm similar to fermion bags, which may solve this problem. As a further approach, we embed the multi-flavour Thirring model in a larger class of four-fermion theories to study the chiral symmetry and its breaking in a wider context.