Open Questions
Research questions I am currently thinking about:
Under what operations is the class closed, which is given by those groups G for which the norm map (in the derived category) of Q[G]-Modules is an equivalence. Both finite groups as well as groups with finitely dominated BG are contained in this class.
Which properties of groups are stable under ultraproducts?
Is the class of groups that satisfy the Farrell-Jones conjecture closed under infinite products?
The Whitehead group of the cyclic group C_n of order n is particularly simple: It is given by a free abelian group of finite rank r, where r is the difference of the number of irreducible real representations and the number of irreducible rational representations. Concrete elements called Bass-Cyclic Units can be given via a formula, however these may fail to generate the entire group. For example for n=5, the element 1-t^2-t^3 in Z[C_5] is the generator of the Whitehead group, but not given by a Bass-Cyclic Unit. Provide concrete formulas of the generators as Elements in the group ring for all n.
Related to this: Can the kernel of the trace map to THH as a subgroup of the Whitehead Group always be completely described?
This question was told to me by Jim Davis: Given an amalgamated product (= pushout) of groups G = K_1 *_H K_2, Waldhausen proved that there is a Mayer-Vietoris-type long exact sequence in K-theory ... -> K_1(Z[K_1])+K_1(Z[K_2]) -> K_1(Z[G])/Nil -> K_0(Z[H]) ->... Is it possible to find an example of G such that the resulting sequence 0 -> cok -> K_1(Z[G])/Nil -> ker -> 0 does not split?
Given a map of spaces X -> Y with finitely dominated fibers, there are wrong way maps in stable homotopy and K-Theory - One can consider the trace map from K-theory to stable homotopy. Does the resulting square commute? (Currently working on this with Marlkiewich, Naef)
Given an E_infty-ring k and an E_1-group G, there should exist a generalization of the Cartan map Perf_k[G] -> Fun(BG, Perf_k) which exists for finite groups. The norm map involving the dualizing spectrum of G of the K-theory of Perf_k should factor through the K-theory of this generalized Cartan map. Similarly, this should hold for all localizing invariants. This will potentially only work in the infinity category of dualizable stable infinity-categories, and potentially needs a finiteness assumption on BG.
By recent work of Ramzi-Sosnilo-Winges, the K-theory functor Cat^perf -> Sp has a section. This implies in particular that Segal K-theory, which group completes the core of a symmetric monoidal category, factors as a functor SymmMonCat -> Cat^Perf. Can this functor be improved in a nice way? Maybe also modeling the classical comparison map K( Proj(R) ) -> K( Perf_R ).
Carlsson-Pederson '95 gave an explicit model for the assembly map in K and L-theory using controlled algebra. Barthels-Efimov-Nikolaus do the same using the stable category of sheaves. There should be a comparison of these two approaches given by a functor "Controlled modules" -> Calk( hatted CoSh( X, R-Mod ) ), which induces an equivalence in K and L-theory.
There should exist a 6-functor formalism on condensed Anima, that encompasses both the 6-functor formalism of locally compact Hausdorff spaces, which sends such a space X to CoSh(X, Sp), as well as the 6-functor formalism of anima, which sends an anima Y to Fun(Y,Sp). The natural functor CoSh(X, Sp) -> Fun(X,Sp) should be incorporated by the assignment of a locally compact Hausdorff space X to its homotopy type (viewed as a condensed anima).
Given an E_1-group G and an analytic ring R, there should be a notion of a "condensed Von-Neumann E_1-algebra" that generalizes the known Von Neumann Algebra in the case when R is the complex numbers. There should be a notion of generalized dimension theory for some notion of module over such algebras. These should give better control over L_2-Betti numbers and related invariants.
The Rosenberg Conjectures. Given a Banach Algebra A over the reals R the two statement should be true:
The algebraic K-theory of A and the topological K-theory agree with mod l coefficients in positive degrees.
The non-positive algebraic K-theory of A is homotopy invariant.
Given a functor F : C -> D, one can define an endofunctor D -> D as the Right Kan Extension of F along itself (if it exists). This is called a codensity monad. One can define "condensed objects" for any codensity-monad. How much of the theory of condensed sets does still work in this context?