arXiv subjects
Publications and source records attributed to Sam Mundy.
We give a new Hecke-theoretic method to control the growth of monodromy in $p$-adic families of automorphic Galois representations for unitary groups and quasisplit classical groups.
We study the residual Eisenstein cohomology of semisimple groups in the context of maximal parabolic subgroups which remain maximal over $\mathbb{R}$. Under certain general hypotheses, we show that these residual representations are cohomological one degree below middle, and one above; however, the classes above middle vanish in the full automorphic cohomology. We offer two proofs of this vanishing, one quick and one more explicit. The latter of the two proofs finds an explicit cochain which provides a primitive to the image of a nontrivial class from the cohomology of such a residual representation. This cochain is valued in regular Eisenstein series. Along the way, we study in detail the archimedean component of the relevant induced representation. In particular, we prove that it has a subrepresentation which is the sum of two discrete series, whose Harish-Chandra parameters we describe, and that the intertwining operator vanishes to order exactly $1$ on that subrepresentation.
We define a certain Hecke theoretic notion of family of smooth admissible representations of $GL_n(F)$, or of products of such groups, where $F$ is a nonarchimedean local field of characteristic zero. While this notion of family is rather weak a priori, we show that it implies strong rigidity properties for the Bernstein-Zelevinsky presentations of the members of such families, as well as for the variation of their epsilon factors (attached to arbitrary functorial lifts). Examples of such families come often from the theory of eigenvarieties, and in this case our results imply analyticity properties for the epsilon factors of the members of these families.
We study the variation of admissible representations of $p$-adic $GL_n$ in families from the point of view of the Bernstein--Zelevinsky classification and show that the ramified parts of these families are rigid. We explain how to apply our results in the context of $GL_n$-eigenvarieties.
Let $F$ be a cuspidal eigenform of even weight and trivial nebentypus, let $p$ be a prime not dividing the level of $F$, and let $\rho_F$ be the $p$-adic Galois representation attached to $F$. Assume that the $L$-function attached to the symmetric cube of $\rho_F$ vanishes to odd order at its central point. Then under some mild hypotheses, and conditional on certain consequences of Arthur's conjectures, we construct a nontrivial element in the Bloch--Kato Selmer group of an appropriate twist of the symmetric cube of $\rho_F$, in accordance with the Bloch--Kato conjectures. Our technique is based on the method of Skinner and Urban. We construct a class in the appropriate Selmer group by $p$-adically deforming Eisenstein series for the exceptional group $G_2$ in a generically cuspidal family and then studying a lattice in the corresponding family of $G_2$-Galois representations. We also make a detailed study of the specific conjectures used and explain how one might try to prove them.
We set up a general framework to compute the exact multiplicity with which certain automorphic representations appear in both the cuspidal and Eisenstein cohomology of locally symmetric spaces. We apply this machinery to Eisenstein series on $GSp_4$ and split $G_2$. In the case of $G_2$, we will also obtain new information about the archimedean components of certain CAP representations using Arthur's conjectures.
In this paper, we give a new proof of an arithmetic analogue of the Riemann-Roch Theorem, due originally to Serge Lang. Lang's result was first proved using the lattice point geometry of Minkowski. By contrast, our proof is completely adelic. It has the conceptual advantage that it uses a different analogue of the Riemann-Roch theorem proved by Tate in his thesis, in a manner similar to the proof of the Riemann-Roch theorem for curves over finite fields which uses Tate's theorem. Thus our proof provides a bridge between a lot of the Riemann-Roch theory that exists in arithmetic.