The classification of irreducible admissible mod \(p\) representations of a \(p\)-adic \(\mathrm{GL}_n\) (Q648060): Difference between revisions

This paper is concerned with the study of the hypothetical mod \(p\) Langlands correspondence. This correspondence should attach each representation \(\mathrm{Gal}(\overline F/F)\to \mathrm{GL}_n(\overline{\mathbb{F}}_p)\) to an admissible smooth representation of \(\mathrm{GL}_n(F)\) over \(\overline{\mathbb{F}}_p.\) Here \(F\) is a finite extension of \(\mathbb{Q}_p,\) and \(\overline{\mathbb{F}_p}\) is the algebraic closure of \(\mathbb{F}_p.\) The paper is inspired by work of \textit{L. Barthel} and \textit{R. Livné} [Duke Math. J. 75, No. 2, 261--292 (1994; Zbl 0826.22019); J. Number Theory 55, No. 1, 1--27 (1995; Zbl 0841.11026)] and generalizes the results of this work to \(GL_n(F).\) Many of the results in this paper hold more generally for a split connected reductive algebraic group over \(F.\) The major results are as follows: {\parindent=6mm\begin{itemize}\item[-] the notion of a ``supersingular representation'' which was defined by Barthel and Livné for \(GL_2,\) is extended to split connected reductive groups; \item[-] an explicit version of the Satake transform is developed, generalizing a formula of Kato-Lusztig for complex representations; \item[-] the irreducibility of generalized Steinberg representations is proved, extending unpublished work of Große-Klönne; all of this is for a general split connected reductive group; \item[-] the classification of admissible smooth mod \(p\) representations of \(\mathrm{GL}_n(F)\) in terms of supersingular data is proved, with the important corollary that ``supersingular'' is equivalent to ``supercuspidal''; \item[-] the submodule structure of parabolically induced representations of \(\mathrm{GL}_n(F)\) is described. \end{itemize}} The classification of admissible smooth representations modulo supercuspidal representations in the mod \(p\) case is similar to the complex case treated by \textit{I. N. Bernstein} and \textit{A. V. Zelevinsky} [Ann. Sci. Éc. Norm. Supér. (4) 10, 441--472 (1977; Zbl 0412.22015); Ann. Sci. Éc. Norm. Supér. (4) 13, 165--210 (1980; Zbl 0441.22014)], with some noteworthy contrasts, pointed out by the author: first, there is no theory of intertwining operators among induced representations. Second, and perhaps relatedly, parabolically induced representations are reducible much more rarely. Finally, when parabolically induced representations are reducible, they are always multiplicity free, in contrast to the complex case.