Notion de θ-r\'egulateurs d'un nombre alg\'ebrique. Conjectures p-adiques
Abstract
Let K/Q be a Galois extension of degree n, of Galois group G, and let η∈ K×. For all large enough prime p, we define, by use of the Frobenius theorem on group determinants, the family (pθ(η) ∈ p)θ of local θ-regulators of η, indexed by the Qp-irreducible characters θ of G. At each pθ (η) is associated a linear representation Lθ δ Vθ, 0 ≤ δ ≤ (1), which characterizes some properties of pθ (η), including its nullity equivalent to δ ≥ 1 (Th. 3.11). When η ∈ × and θ = 1, p1 (η) is the p-Fermat quotient of η. When η is a "Minkowski unit", each pθ (η), θ 1, gives the residue modulo p of the θ-component of p1-n Regp (K), where Reg(K) is the classical p-adic regulator of K. We suggest that the "probability" of (pθ(η) = 0 and Lθ δ Vθ) is O(1)pf δ2, where f is a suitable residue degree of p. We conjecture that p1-n Regp(K), which measures the order of the p-torsion group in Abelian p-ramification over K, is for p large enough a p-adic unit except perhaps for a set of prime numbers of zero density. For these cases said "of minimal p-divisibility" (Def. 3.17), it remains possible, η being then a "partial local pth power" at p, to propose, in connection with the ABC conjecture, a stronger conjecture leading to the same conclusion for all large enough p (Section 7). Some other conjectural aspects on the Fermat quotient are discussed. We precise and verify these properties through numerical studies on various fields and publish the corresponding "PARI" programs.
Turn this paper into a lesson
ArcXiv compiles a structured reading guide from this paper's metadata: plain-English importance, contributions, prerequisite concepts, which sections to read first, flashcards, and a quiz. Grounded in the abstract, never invented.