Equivariant pretheories and invariants of torsors (Q692786): Difference between revisions

The paper under review develops a notion of equivariant pretheories and equivariant homology of a Rost cycle module and applies these concepts to the study of \(G\)-torsors, where \(G\) is an algebraic group acting on a quasi-projective variety. The idea underlying these constructions is roughly speaking an amalgamation of two ideas which have previously each for themselves been a success story. On the one hand, there is Rost's theory of cycle modules [\textit{M. Rost}, Doc. Math., J. DMV 1, 319--393 (1996; Zbl 0864.14002)]. This is a quite general formalism of cycle complexes that encompasses a great variety of examples. It provides a uniform setup establishing all basic properties one desires from classical cohomology theories (say localization sequences, \(\mathbb{A}^1\) homotopy invariance, transfers, etc.) from some basic datum on points/fields. On the other hand, there is \textit{D. Edidin} and \textit{W. Graham}'s theory of equivariant Chow groups [Invent. Math. 131, No. 3, 595--644 (1998; Zbl 0940.14003)]. This theory is inspired from the construction of equivariant cohomology in algebraic topology. As in topology one encounters the problem that interesting actions have the bad habit not to be free. To resolve this an algebraic replacement for the classifying space of \(G\) is built, but a bit different than in topology, by finding suitable representations containing (viewed as affine space) a suitable open on which \(G\) acts freely. This goes back to an idea of Totaro. The authors now combine these two theories, leading to their concept of \(G\)-equivariant cycle homology of a Rost cycle module. They show that this theory has all properties (e.g. pullbacks, pushforwards) one can and should reasonably ask for. The authors give a large number of examples illustrating this notion. Moreover, they develop a spectral sequence modelled after Merkurjev's spectral sequence for \(G\)-actions in equivariant algebraic \(K\)-theory and a theorem of \textit{N. A. Karpenko} and \textit{A. S. Merkurjev} in their setup (established in [Adv. Math. 205, No. 2, 410--433 (2006; Zbl 1119.14041)]). Again, many examples are given.