Perturbation Theory around Non-Nested Fermi Surfaces I. Keeping the Fermi Surface Fixed
Joel Feldman (UBC, Vancouver), Manfred Salmhofer (ETH, Zurich) and, Eugene Trubowitz (ETH, Zurich)

TL;DR
This paper develops a renormalized perturbation theory for many-fermion systems with non-nested, non-spherical Fermi surfaces, establishing convergence, differentiability, and graph classification results, with applications to the Hubbard model.
Contribution
It introduces a comprehensive renormalization approach for non-nested Fermi surfaces, including graph classification and power counting, extending the analysis to the Hubbard model away from half-filling.
Findings
Counterterms converge to a finite, differentiable limit.
The map from renormalized to bare band structure is locally injective.
Ladder graphs are the primary source of factorial divergences in perturbation series.
Abstract
The perturbation expansion for a general class of many-fermion systems with a non-nested, non-spherical Fermi surface is renormalized to all orders. In the limit as the infrared cutoff is removed, the counterterms converge to a finite limit which is differentiable in the band structure. The map from the renormalized to the bare band structure is shown to be locally injective. A new classification of graphs as overlapping or non-overlapping is given, and improved power counting bounds are derived from it. They imply that the only subgraphs that can generate factorials in the order of the renormalized perturbation series are indeed the ladder graphs and thus give a precise sense to the statement that `ladders are the most divergent diagrams'. Our results apply directly to the Hubbard model at any filling except for half-filling. The half-filled Hubbard model is treated in…
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
