Stable representation homology and Koszul duality

Abstract

This paper is a sequel to [BKR], where we studied the derived affine scheme DRepn(A) of the classical representation scheme Repn(A) for an associative k-algebra A. In [BKR], we have constructed canonical trace maps Trn(A): HC(A) -> H[DRepn(A)]GL extending the usual characters of representations to higher cyclic homology. This raises a question whether a well known theorem of Procesi [P] holds in the derived setting: namely, is the algebra homomorphism Sym[Trn(A)]: Sym[HC(A)] -> H[DRepn(A)]GL defined by Trn(A) surjective ? In the present paper, we answer this question for augmented algebras. Given such an algebra, we construct a canonical dense DG subalgebra DRep∞(A)Tr of the topological DG algebra DRep∞(A)GL∞. It turns out that on passing to the inverse limit (as n -> ∞), the family of maps Sym[Trn(A)] "stabilizes" to an isomorphism Sym[HC(A)] = H[DRep∞(A)Tr]. The derived version of Procesi's theorem does therefore hold in the limit. However, for a fixed (finite) n, there exist homological obstructions to the surjectivity of Sym[Trn(A)], and we show on simple examples that these obstructions do not vanish in general. We compare our result with the classical theorem of Loday-Quillen and Tsygan on stable homology of matrix Lie algebras. We show that the relative Chevalley-Eilenberg complex C(gl∞(A), gl∞(k); k) equipped with the natural coalgebra structure is Koszul dual to the DG algebra DRep∞(A)Tr. We also extend our main results to bigraded DG algebras, in which case we show that DRep∞(A)Tr = DRep∞(A)GL∞. As an application, we compute the (bigraded) Euler characteristics of DRep∞(A)GL∞ and HC(A) and derive some interesting combinatorial identities.

0

Turn this paper into a lesson

ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.

Discussion (0)

Sign in to join the discussion.

Loading comments…