Successive-maxima conjecture for non-genus class numbers of abelian fields

Let GG be an abelian group and let KGs{\mathcal K}_G^s be the family of abelian fields with Galois group GG and signature ss. For a conductor ff, let FG,fs{\mathcal F}_{G,f}^s consist of fields in this family with conductor ff, and let FG,fs{\mathcal F}'{}_{G,f}^s consist of fields with conductor dividing ff. For KK in the family, write DKD_K for its discriminant, gK/Qg_{K/\mathbb Q} for its genus number, and ClK\operatorname{Cl}_K for its class group. For 0<ε<10<\varepsilon<1, define

Cε(FG,fs)=[KFG,fsClKgK/Q(DK)ε]1/FG,fs,C_{\varepsilon}({\mathcal F}_{G,f}^s)=\left[\prod_{K\in{\mathcal F}_{G,f}^s}\frac{|\operatorname{Cl}_K|}{g_{K/\mathbb Q}(\sqrt{|D_K|})^{\varepsilon}}\right]^{1/|{\mathcal F}_{G,f}^s|},

with analogous definitions of Cε(FG,fs)C'_{\varepsilon}({\mathcal F}'{}_{G,f}^s) and Cε(K)C_{\varepsilon}(K). Let r(G)r(G) be the minimal number of generators of GG. If G=i=1tpini|G|=\prod_{i=1}^t p_i^{n_i}, put R(G)=i=1tniR(G)=\sum_{i=1}^t n_i.

Successive-maxima conjecture. (i) As ff increases from its minimum, the successive maxima of Cε(FG,fs)C_{\varepsilon}({\mathcal F}_{G,f}^s), Cε(FG,fs)C'_{\varepsilon}({\mathcal F}'{}_{G,f}^s), and Cε(K)C_{\varepsilon}(K) occur, with probability 11, for conductors satisfying

r(G)ω(f)R(G).r(G)\leq\omega(f)\leq R(G).

This property is independent of ε\varepsilon. (ii) For every 0<ε<10<\varepsilon<1, each of these sequences of successive maxima is infinite.

The probability-one qualification means that the assertion may fail on a set of conductors of density zero. The statement is motivated by computations of class numbers after removing the genus contribution; the behavior of the number of prime divisors in the indicated interval remains unclear, and the possible even-order wild-ramification exception is explicitly neglected in the source.

Sources & referencesView supporting material

Primary source

Georges Gras, “Successive maxima of the non-genus part of class numbers”, arXiv:1911.13115 (2019).

Progress summary

Never refreshed

Nothing recorded yet. Refresh searches the literature and the public web for attempts on this problem, and writes the first summary here.

Solutions 0

No solutions have been posted yet.