
ALT We prove a conjecture by J. Zhao which states that every finite multiple zeta value associated with a nonempty index is a ℚ-linear combination of ζ_A(k₁,…, kᵣ, 2, 1), where kᵢ∈ {2, 3}. Our proof proceeds by constructing a basis for the space of formal multiple zeta values defined by the confluence relations. This approach also reduces the Kaneko–Zagier conjecture (apart from compatibility with the depth filtration) to linear independence conjectures for ordinary and finite multiple zeta values, without using motivic theory.

ALT A Borwein polynomial has coefficients in -1,0,1 and nonzero constant and leading coefficients. We seek Borwein polynomials having a multiple root outside the unit disk. To this end, we consider the powers L² and L³ of Lehmer's polynomial L, which has a real root τ>1. Since L has exceptionally small Mahler measure, the theorem of Bombieri and Vaaler guarantees the existence of Borwein multiples of these powers. Such multiples have τ as a root of multiplicity at least 2 and 3, respectively. Mossinghoff (2003, Algorithm 3, p. 323) constructed examples that, under the substitution x↦ -x, yield such multiples. The main contribution of this paper is to determine the least possible degrees of Borwein multiples of L² and L³, with and without symmetry restrictions on their coefficients, and to describe their degree spectra. For L², the minimum degree is 34 without symmetry restrictions, 38 among self-reciprocal multiples, and 41 among antireciprocal multiples. For L³, the corresponding minimum

ALT Using the second moment of Dirichlet L-functions on the critical line over the major arcs 𝔐(mathscr Lᵇ), with τ=y³x⁻¹mathscr L⁻ᵇ₁, and excluding a small neighborhood of the centers of these arcs, i.e., those α satisfying |α-ᵃ⁄q|>(8πy²)⁻¹, for y≥ x¹⁻ʳⁱᵖᵗˢᵗʸˡᵉ¹/ʳⁱᵖᵗˢᵗʸˡᵉ⁹⁻⁴√²mathscr Lᶜ₂, where c₂=2A+24+(√2-1)b₁/2√2-1, we obtain the estimate S₂(α;x,y)=∑ₓ₋y<ₙ≤ₓΛ(n)e(αn²)≪ y mathscr L⁻ᴬ. Moreover, in a small neighborhood of the center of the major arcs, defined by |α-ᵃ⁄q|≤(8πy²)⁻¹, for y≥ xʳⁱᵖᵗˢᵗʸˡᵉ⁵⁄ᵣᵢₚₜₛₜʸₗₑ₈mathscr L¹⁵ᴬ⁺⁰²⁵ᵇ⁺¹⁸, an asymptotic formula with a remainder term is obtained for S₂(α;x,y), where A, b₁, and b are arbitrary fixed positive constants, and mathscr L=ln xq.

ALT We establish Kim's conjectural dimension inequality for smooth projective geometrically connected curves C/ℚ of genus at least two with a rational basepoint. For every odd prime p of good reduction, we prove that dim Hf¹(Gℚ,Σ,Uₙét) < dim(Uₙᵈᴿ/F⁰) for all sufficiently large n, where Uₙét and UₙᵈR are the n'th lower central series truncations of the ℚₚ-pro-unipotent geometric étale, and de Rham fundamental groups of C, and Σ is any finite set of places containing p, ∞, and the primes of bad reduction. Inspired by Coates and Kim's work on curves with CM Jacobians, we use noncommutative Iwasawa theory to bound second Galois cohomology through dimensions of kernels of certain Iwasawa operators, acting on finite dimensional representations of the ambient generalized symplectic group. We establish bounds for these kernels in terms of the highest weight and dimension of the ambient symplectic representation. No restriction is imposed on the Jacobian's geometric endomorphism ring or its Tate-mo

ALT For a prime p, we consider two important families of integral quadratic lattices associated to isomorphism classes of supersingular elliptic curves over \̅mathbbFₚ, the lattices of endomorphisms of trace zero, and the Gross lattices of endomorphisms. We prove effective equidistribution of the shapes of these lattices as p→∞ through primes, in the arithmetic locally symmetric space of unimodular lattices up to isometry. As an application, we obtain, as p→∞, the asymptotic distribution of successive minima of these lattices. These results are proved using tools from homogeneous dynamics, in particular, the key dynamical input is a recent adelic equidistribution theorem due to Einsiedler, Margulis, Mohammadi, and Venkatesh.

ALT Andrews and Newman introduced the minimal excludant (mex) of an integer partition, and Aricheta and Donato extended it to overpartitions. More recently, Meher and Tejash, Bhowmick and Meher studied the analogous statistic for ℓ-regular partitions. In this article, for an integer ℓ≥ 2, we define the minimal excludant omexℓ(π) of an ℓ-regular overpartition π as the smallest positive integer that is not divisible by ℓ and does not occur as a non-overlined part of π. We study ℓ(n), the sum of omexℓ(π) over all ℓ-regular overpartitions π of n. We obtain generating functions for ℓ(n) as eta-quotients for ℓ=2 and ℓ=3. Later, motivated from, we obtained explicit generating functions for ℓ(n). We show that ℓ(n) is weakly increasing in n, and we derive Hardy–Ramanujan type asymptotic formulas for 2(n) and 3(n). We then determine the parity of both functions completely: 2(n) is even for every n≥ 1, while 3(n) is odd if and only if n=0 or n is a centred hexagonal number. Also, we obtain for every

ALT We obtain lower bounds for the ranks of universal forms over totally real cubic fields using techniques from geometry-of-numbers and the parametrisation of cubic rings by binary cubics due to Delone–Faddeev. Additionally, our methods give lower bounds for totally real primitive fields of any degree with codifferent generated by a totally positive element.
![Let Pᵤ(X)=XQ(X)(X-u)+1, where Q(X)∈ ℤ[X] is monic, separabel, has constant term ± 1, and degree q≥ 1. Denote (Aₙ(u)) the linear recursive sequence having characteristic polynomial Pᵤ(X) and initial terms A₀,…,Aq₊₁∈ ℤ such that at least one of them is non-zero. Then there exist effectively computable constants n₀,u₀ depending only on Q(X) and max{|A₀|,…, |Aq₊₁|} such that if |Aₙ(u)|=|Aₘ(u)|holds for some integers u,n,m with u>u₀ and n≠m then max{](https://pbs.twimg.com/media/HXPZSKSXcAAhM-y.png?name=small&format=webp)
ALT Let Pᵤ(X)=XQ(X)(X-u)+1, where Q(X)∈ ℤ[X] is monic, separabel, has constant term ± 1, and degree q≥ 1. Denote (Aₙ(u)) the linear recursive sequence having characteristic polynomial Pᵤ(X) and initial terms A₀,…,Aq₊₁∈ ℤ such that at least one of them is non-zero. Then there exist effectively computable constants n₀,u₀ depending only on Q(X) and max{|A₀|,…, |Aq₊₁|} such that if |Aₙ(u)|=|Aₘ(u)|holds for some integers u,n,m with u>u₀ and n≠m then max{

ALT We study non-vanishing of central standard L-values of spherical cuspidal GL₃ newforms of prime level. A logarithmic Möbius mollifier with finite inverse Euler factors, combined with a conjugate term from the functional equation, gives a non-vanishing harmonic proportion of at least 1/2-ε for every ε>0 and all sufficiently large prime levels. The required mixed moment is evaluated by an asymmetric approximate functional equation.

ALT We study rational lines on smooth intersections of two quadrics and on smooth cubic hypersurfaces over number fields. For an intersection of two quadrics in n variables, we prove the Hasse principle and weak approximation for the variety of lines when n≥ 11, and more generally for the variety of r-planes when n≥ 3r+8. In this range an r-plane exists over the ground field if and only if one exists over every real completion. For the variety of lines, we also prove the Hasse principle and weak approximation when n≥ 8, conditional on the finiteness of certain Tate-Shafarevich groups. For a smooth cubic hypersurface in n variables, we prove that a line exists over the ground field when n≥ 23.

ALT We make and test the conjecture that for p-adic L-functions associated to quadratic Dirichlet characters and their Teichmüller twists, all local statistics of zeros in the unit disk match those of p-adic random matrix eigenvalues. This generalizes the case of the number of zeros in the unit disk previously conjectured by Ellenberg-Jain-Venkatesh, which was well-supported by numerics, as well as the case of the discriminant, where their numerics were more ambiguous. Using recent random matrix results on local statistics of eigenvalues, we make the random matrix prediction explicit for many statistics, analogous to local densities and pair-correlations in the complex setting. To test these, we compute data on p-adic L-functions for characters of conductor up to 10⁷, and for certain statistics in windows above 10¹⁰. We find good agreement with the new random matrix predictions for these statistics, as well as for the discriminant.

ALT Let K/𝐐ₚ be unramified, with p>2. For a 1-motive with constant lattice, split torus, and good abelian quotient, we compute the semistable extension complex from toric valuations and formal-group logarithms. A canonical monodromy boundary exists for every semistable representation; it detects splitting over crystalline periods and the increase in monodromy rank. For the stated 1-motives, this boundary and the unramified quotient recover the restricted local Tate pairing, with exponential radicals. Nonsplit elliptic examples, Raynaud uniformization, and a height-one positive-period bound explain the geometric range. For toric motives we construct actual periods, convergent Robba bases, and integral action matrices. A toric basis family has perfect projective/trace-zero matrix duality, including its exact integral defect. An explicit residue-zero solution operator constructs every dual Herr class. Applied componentwise, the same operator gives relative full dual lifts for general constan

ALT We count ordered commuting pairs in a full lattice Λ in a reductive Lie algebra \g over mathbbQ. For a fixed convex body B containing the origin in its interior, the count in HB is CΛ,BH^dim\g+dimZ(\g) with a power-saving error. The leading term comes from pairs whose projections to each simple ideal are linearly dependent over its centroid. Their slopes index an absolutely convergent sum of section volumes for CΛ,B. Subtracting the contribution of dependent pairs from the finite-field Fourier sums gives cancellation at generic dependent frequencies. For independent frequencies, we bound the dimension of the locus where the corresponding hyperplane contains an irreducible component of a nonzero bracket fibre. For n≥3, the number of ordered commuting pairs in M_n(mathbbZ) with entries of absolute value at most H is 2(n+3)ζ(n²-2)/(n+1)ζ(n²-1) (2⌊H⌋+1)^n²+1 +O_n(H^n²+1-1/(n+1)), confirming the exponent n²+1 predicted by Browning, Sawin and Wang.

ALT We prove that all mod 2 Galois representations of a global field lift modulo 4. In contrast, for every odd prime p, we exhibit examples of global Galois representations modulo p that do not lift modulo p². Our methods also allow us to prove liftability of generalized quaternion groups and hence to complete the classification of liftable finite groups initiated by Khare and Merkurjev.

ALT Let f=E₁(χ₁,χ₂) be an irregular Eisenstein series of weight one, that is, with χ₁(p)=χ₂(p), and consider the generically cuspidal ordinary family passing through it. When χ₁, χ₂ and χ₁χ₂⁻¹ are primitive of the same conductor, we show that the specialization at f of the modular-symbol p-adic L-function of this family, with periods taken in the cohomology of the compactified modular curves, factors as a product of two Kubota–Leopoldt p-adic L-functions attached to χ₁ and χ₂. Since there are no complex critical values in weight one, the factorization is obtained directly from the cohomological construction: we compute the tame-twisted modular measure by means of the localized Sharifi map and explicit reciprocity, and show that tame twisting identifies the compact period modules at weight one. This gives a weight-one analogue of the factorization of Bellaïche–Dasgupta for critical-slope Eisenstein series. As applications, and under explicit specialization hypotheses, we compare Kato classe
![We study the multidimensional polynomial Szemerédi theorem over prime fields. Let P₁,…,Pₘ∈\Fₚ[t]ᵏ be block independent vector polynomials of degree at most d, with Pᵢ(0)=0 for every i. We prove that there is a constant c=c(k,m,d)>0 such that, for every sufficiently large prime p, every set A⊂\Fₚᵏ with |A|≥ pᵏ⁻ᶜ contains a configuration x,x+P₁(t),…,x+Pₘ(t) with t∈\Fₚ∖{0}.](https://pbs.twimg.com/media/HXD4ra4W8AAolV8.png?name=small&format=webp)
ALT We study the multidimensional polynomial Szemerédi theorem over prime fields. Let P₁,…,Pₘ∈\Fₚ[t]ᵏ be block independent vector polynomials of degree at most d, with Pᵢ(0)=0 for every i. We prove that there is a constant c=c(k,m,d)>0 such that, for every sufficiently large prime p, every set A⊂\Fₚᵏ with |A|≥ pᵏ⁻ᶜ contains a configuration x,x+P₁(t),…,x+Pₘ(t) with t∈\Fₚ∖{0}.

ALT We give a Diophantine definition of the integers in the rational numbers conditional on the abc conjecture, Dickson's conjecture, and finiteness of 2-primary parts of Shafarevich–Tate groups. We also discuss some consequences of the existence of a Diophantine definition of the integers in the rational numbers. The goal of this work is to serve as an entry point to recent developments around Hilbert's tenth problem for the rational numbers.

ALT Let q be a prime power and let \ell be a divisor of q-1. We classify the permutation polynomials of Fq of polynomial index exactly \ell up to quasi-multiplicative equivalence G(X)≡ uF(vX^d)±odX^q-X, where u,v∈Fq* and \gcd(d,q-1)=1, and we count the classes. For polynomials vanishing at 0, the classes are the orbits of an explicit finite group on vectors of values at the \ell-th roots of unity. Through the permutation it induces on the roots of unity, each orbit lies over a double coset of the symmetric group S_\ell with respect to its translation subgroup and an explicit affine group, and the stabilizers are explicit. For polynomials with nonzero constant term the action is free. Burnside's lemma, Möbius inversion over the coefficient support and congruences along the cycles of affine permutations give an exact formula for the number of classes. For fixed \ell≥slant2 these numbers are quasi-polynomials in s=(q-1)/\ell of degree at most \ell-1, with period dividing \ell for odd \ell and

ALT We introduce lower-truncated p-adic multiple zeta-star values and construct their harmonic regularizations. We show that they are independent of the branch of the p-adic logarithm and satisfy harmonic product relations. We express finite multiple harmonic sums as alternating convolutions of harmonic-regularized p-adic multiple zeta values and lower-truncated p-adic multiple zeta-star values. This gives a p-adic analogue of the classical truncation identity, which reduces to the inversion relation between ordinary and star harmonic-regularized values when the truncation point is 1. For truncation point p, it relates finite multiple zeta values to p-adic multiple zeta values.

