One-dimensional Chern-Simons theory and the \(\widehat{A}\) genus (Q740543)

This paper constitutes a mathematically rigorous construction of a quantum field theory via the Batalin-Vilkovisky formalism. More specifically, the authors construct a one dimensional Chern-Simons theory which depends on the choice of an \(L_\infty\)-algebra. Then, using formal geometry to obtain an \(L_\infty\)-algebra from a smooth manifold \(M\), they interpret their Chern-Simons theory as a non-linear sigma model for the loop space of the cotangent bundle of \(M\). The main theorem is an identification of the homotopy type of the quantum observables in terms of well-understood topological and geometric invariants. Roughly, their observables are some variation of the de Rham forms of the cotangent bundle that is deformed by the natural symplectic structure of the cotangent bundle and a version of the \(\hat{A}\) genus of the underlying \(L_\infty\)-algebra or manifold. The approach is that of Kevin Costello and his school and this paper fits into a larger body of literature that employ the same language and axiomatics. In this formalization, a quantum field theory is a family of algebraic structures for varying length scale with the correct limiting behavior when the length scale becomes small. There should be a relation (renormalization group flow) between the structures at different length scales, which is a kind of transfer or homological perturbation using Feynman diagrams. In order to construct their theory, the authors begin with a length-scale zero classical field theory and run an a priori illegal version of this flow to generate the algebraic structures at arbitrary length scale. It is a feature of the particular kind of Chern-Simons theory that they are considering that diagrams with more than one loop do not contribute. Then they are able to calculate that this illegal flow yields a well-defined family of algebraic structures satisfying all of the requisite properties. In order to understand this quantum theory homotopically, they take the limit as the length scale becomes large. This yields a well-defined algebraic structure equivalent to those at finite length scale, but whose structure is simpler and more canonical. More specifically, the zero-loop part is essentially the de Rham forms of the cotangent bundle and the one-loop part can be written explicitly in terms of characteristic classes of the classifying space of the coefficient \(L_\infty\)-algebra. When the \(L_\infty\)-algebra arises from their formal geometric constructions on a smooth manifold, these classes can be identified with characteristic classes of the manifold itself. There is quite a bit of machinery of diverse sorts that enters into the background, motivation, constructions, and proofs. However, the paper is carefully organized so that each section uses one main tool and is more or less independent of the other sections. There are also numerous expository breaks to help orient the reader, discuss the context of relevant literature, and lay the groundwork for the rigorous constructions to follow.