We study deformation rings of an n-dimensional representation $\overline{\rho}$, defined over a finite field of characteristic $\ell$, of the arithmetic fundamental group $\pi_1(X)$, where X is a geometrically irreducible, smooth curve over a finite field k of characteristic p ($ \neq \ell$). When $\overline{\rho}$ has large image, we are able to show that the resulting rings are finite flat over $\mathbf{Z}_\ell$. The proof principally uses a Galois-theoretic lifting result of the authors in Part I of this two-part work, a lifting result for cuspidal mod $\ell$ forms of Ogilvie, Taylor–Wiles systems and the result of Lafforgue. This implies a conjecture of de Jong for representations of $\pi_1(X)$ with coefficients in power series rings over finite fields of characteristic $\ell$, that have this mod $\ell$ representation $\overline{\rho}$ as their reduction. A proof of all cases of the conjecture for $\ell>2$ follows from a result announced by Gaitsgory. The methods are different.

We give a purity result for two kinds of exponential sums of the type $\sum_{x\in k^n}\psi(f(x))$, where k is a finite field of characteristic p and $\psi:k\to{\mathbb C}^\star$ is a non-trivial additive character. In the first case, $f\in k[x_1,\dots,x_n]$ is a polynomial of degree divisible by p whose highest-degree homogeneous form defines a non-singular projective hypersurface, and in the second case, f is a polynomial of degree prime to p whose highest-degree homogeneous form defines a projective hypersurface with isolated singularities.

Derivation and extended double shuffle (EDS) relations for multiple zeta values (MZVs) are proved. Related algebraic structures of MZVs, as well as a ‘linearized’ version of EDS relations are also studied.

We construct families of hyperelliptic curves over ${\mathbb Q}$ of arbitrary genus g with (at least) g integral elements in K2. We also verify the Beilinson conjectures about K2 numerically for several curves with g = 2, 3, 4 and 5. The first few sections of the paper also provide an elementary introduction to the Beilinson conjectures for K2 of curves.

We construct a map between Bloch's higher Chow groups and Deligne homology for smooth, complex quasiprojective varieties on the level of complexes. For complex projective varieties this results in a formula which generalizes at the same time the classical Griffiths Abel–Jacobi map and the Borel/Beilinson/Goncharov regulator type maps.

Let R be the exterior algebra in n + 1 variables over a field K. We study the Auslander–Reiten quiver of the category of linear R-modules, and of certain subcategories of the category of coherent sheaves over Pn. If n > 1, we prove that up to shift, all but one of the connected components of these Auslander–Reiten quivers are translation subquivers of a ${\bf Z} A_{\infty}$-type quiver. We also study locally free sheaves over the projective n-space Pn for n > 1 and we show that each connected component contains at most one indecomposable locally free sheaf of rank less than n. Finally, using results from the theory of finite-dimensional algebras, we construct a family of indecomposable locally free sheaves of arbitrary large ranks, where the ranks can be computed using the Chebysheff polynomials of the second kind.

Let X be a smooth complex projective surface and let C(X) denote the field of rational functions on X. In this paper, we prove that for any m > M(X), there exists a rational dominant map $f \colon X \to Y$, which is generically finite of degree m, into a complex rational ruled surface Y, whose monodromy is the alternating group Am. This gives a finite algebraic extension C(X): C(x1, x2) of degree m, whose normal closure has Galois group Am.

We classify all Gorenstein Fano threefolds with at worst canonical singularities for which the anticanonical system |−K| has a nonempty base locus.

Using limit linear series and a result controlling degeneration from separable maps to inseparable maps, we give a formula for the number of rational functions (up to automorphism of the target) on the projective line with ramification to order ei at general points Pi, in the case that all ei are less than the characteristic. Unlike the case of characteristic 0, the answer is not given by Schubert calculus, nor is the number of maps always finite for distinct Pi, even in the tamely ramified case. However, finiteness for general Pi, obtained by exploiting the relationship to branched covers, is a key part of the argument.

Suppose that R is a two-dimensional normal standard-graded domain over a finite field. We prove that there exists a uniform Frobenius test exponent b for the class of homogeneous ideals in R generated by at most n elements. This means that for every ideal I in this class we have that $f^{p^b}\in I^{[p^b]}$ if and only if $f\in I^{\rm F}$. This gives in particular a finite test for the Frobenius closure. On the other hand we show that there is no uniform bound for Frobenius test exponent for all homogeneous ideals independent of the number of generators. Under similar assumptions we prove also the existence of a bound for tight closure test ideal exponents for ideals generated by at most n elements.

We show that a linear subspace of a reductive Lie algebra $\operatorname{\mathfrak g}$ that consists of nilpotent elements has dimension at most $\frac{1}{2}(\dim\operatorname{\mathfrak g}-\operatorname{rk}\operatorname{\mathfrak g})$, and that any nilpotent subspace attaining this upper bound is equal to the nilradical of a Borel subalgebra of $\operatorname{\mathfrak g}$. This generalizes a classical theorem of Gerstenhaber, which states this fact for the algebra of $(n\times n)$-matrices.

We compute the algebras of self-extensions of the vacuum module and the Verma modules over an affine Kac–Moody algebra $\hat{\mathfrak g}$ in suitable categories of Harish-Chandra modules. We show that at the critical level these algebras are isomorphic to the algebras of differential forms on various spaces of opers associated to the Langlands dual Lie algebra of ${\mathfrak g}$, whereas away from the critical level they become trivial. These results rely on and generalize the description of the corresponding algebras of endomorphisms obtained by Feigin and Frenkel and the description of the corresponding graded versions due to Fishel, Grojnowski and Teleman.

If G is a (split) Kac–Moody group over a field K endowed with a real valuation $\omega$, we build an action of G on a geometric object $\mathcal I$. This object is called a building, as it is an union of apartments, with the classical properties of systems of apartments. However, these apartments are more exotic: that associated to a torus T may be seen as the gluing of all Satake compactifications of affine apartments of T with respect to spherical parabolic subgroups of G containing T. Another geometric realization of these apartments makes them look more like the apartments of $\Lambda$-buildings; then the translations of the Weyl group act only on infinitely small elements of the apartment, so we call these buildings microaffine.

Let d be the degree of an algebraic one-dimensional foliation $\mathcal F$ on the complex projective space ${\mathbb P}_n$ (i.e. the degree of the variety of tangencies of the foliation with a generic hyperplane). Let $\Gamma$ be an algebraic solution of degree $\delta$, and geometrical genus g. We prove, in particular, the inequality $(d-1)\delta+2-2g\geq {\mathcal B}(\Gamma)$, where ${\mathcal B}(\Gamma)$ denotes the total number of locally irreducible branches through singular points of $\Gamma$ when $\Gamma$ has singularities, and ${\mathcal B}(\Gamma)=1$ (instead of 0) when $\Gamma$ is smooth. Equivalently, when $\Gamma=\bigcap_{\lambda=1}^{n-1} S_\lambda$ is the complete intersection of n - 1 algebraic hypersurfaces $S_\lambda$, we get $(d+n-\sum_{\lambda=1}^{n-1}\delta_\lambda)\delta \geq {\mathcal B}(\Gamma)-{\mathcal E}(\Gamma)$, where $\delta_\lambda$ denotes the degree of $S_\lambda$ and ${\mathcal E}(\Gamma)=2-2g+(\sum_\lambda\delta_\lambda-(n+1))\delta$ the correction term in the genus formula. These results are also refined when $\Gamma$ is reducible.