Local unitary cocycles of \(E_0\)-semigroups (Q1011437)

\textit{B.\,Tsirelson} [New York J.~Math.\ 14, 539--576 (2008; Zbl 1177.46048)] showed that Arveson systems need not be \textit{amenable}. That is, the automorphism group of a type II Arveson system need not act transitively on the set of normalized units of that Arveson system. (Note that in the present paper, the statement would be phrased as follows: The group of local unitary cocycles with respect to a spatial \(E_0\)-semigroup need not act transitively on the set of units of that semigroup.) The authors investigate a closely related problem; in the abstract, they write: ``This paper concerns the structure of the group of local unitary cocycles, also called the gauge group, of an \(E_{0}\)-semigroup. The gauge group of a spatial \(E_{0}\)-semigroup has a natural action on the set of units by operator multiplication. Arveson has characterized completely the gauge group of \(E_{0}\)-semigroups of type I, and as a consequence it is known that in this case the gauge group action is transitive. In fact, if the semigroup has index \(k\), then the gauge group action is transitive on the set of \((k+1)\)-tuples of appropriately normalized independent units. An action of the gauge group having this property is called \((k+1)\)-fold transitive. We construct examples of \(E_{0}\)-semigroups of type II and index 1 which are not 2-fold transitive. These new examples also illustrate that an \(E_{0}\)-semigroup of type II\(_k\) need not be a tensor product of an \(E_{0}\)-semigroup of type II\(_{0}\) and another of type I\(_k\).'' I would like two make some comments. Tsirelson's example for a non-amenable type II\(_1\) system is \textit{a~fortiori} an example for a non-\(2\)-fold transitive type II\(_1\) system. The statement that type II\(_1\) systems need not be tensor products of a type II\(_0\) system and a type I\(_1\) system has already been proved as Proposition 4.32 in [\textit{V.\,Liebscher}, Random sets and invariants for (type III) continuous tensor product systems of Hilbert spaces (Mem.\ Am.\ Math.\ Soc.\ 930) (2009; Zbl 1190.46053)], a paper that has been around as an arXiv preprint since 2003. Although I could guess what \((k+1)\)-fold transitive means, it would have been nice to see a slightly more explicit definition.