Semibounded unitary representations of double extensions of Hilbert-loop groups (Q486825): Difference between revisions

The paper under review brings tremendous progress in representation theory of infinite-dimensional Lie groups. A wealth of information and interesting new results on Fréchet-Lie groups are collected in the appendices. (See particularly the infinitesimal characterization of analytic operator-valued functions on Lie groups modeled on Fréchet spaces, whose multiplication is locally given by the Baker-Campbell-Hausdorff series; in particular, this applies to all Lie groups modeled on Banach spaces.) However, the main body of the paper focuses on groups \(\widehat{\mathcal L}_\phi(K)\) which are obtained by double extension from twisted loop groups \({\mathcal L}_\phi(K)\), where \(\phi\) is an automorphism of finite order of \(K\), while \(K\) is a simple Lie group modeled on a real Hilbert space such that the corresponding scalar product on the Lie algebra of \(K\) is invariant under the adjoint action. These are the groups referred to in the title of the paper, and they are natural infinite-rank versions of affine Kac-Moody groups. In fact, this is the first time when global aspects of such Lie groups are investigated. On the other hand, a semibounded unitary representation of any Lie group \(G\) is any unitary representation \(\pi\) of \(G\) whose corresponding infinitesimal generators \(\text{id}\pi(x)\) are uniformly bounded from above for \(x\) in some nonempty open subset of the Lie algebra of \(G\). This is a quite natural condition which is satisfied by many representations that occur in various areas of mathematics and in physics. With this terminology at hand, the main result is an explicit classification of the irreducible semibounded unitary representations of the above infinite-dimensional Lie groups \(\widehat{\mathcal L}_\phi(K)\). To this end, many interesting techniques must be developed or adapted, notably the method of holomorphic induction in an appropriate infinite-dimensional setting, which provides a method of constructing semibounded representations. Although the emphasis is on infinite dimensions, new results are obtained here even in the case when the above \(K\) is a compact Lie group, for instance the fact that the so-called positive energy representations are semibounded. The paper under review is very well written and also contains a generous discussion on open problems and perspectives of further development.