Asymptotic persistence density conjecture for Benford sequences

From papers

Let bb range over integer bases greater than 11. A base is B-EQUI-PERS at sample size NN when there exists a continued-fraction convergent denominator qkNq_k\leq N such that ak+1qk>NlogNa_{k+1}q_k>N\log N, where ak+1a_{k+1} is the next partial quotient of the continued fraction of log10(b)\log_{10}(b). Asymptotic persistence density conjecture. The persistence density among integer bases converges:

ρ=limB#{bB:b is B-EQUI-PERS}B1=112.\rho_\infty=\lim_{B\to\infty}\frac{\#\{b\leq B:b\text{ is B-EQUI-PERS}\}}{B-1}=\frac{1}{12}.

Extended computation gives densities close to 1/121/12, while a Gauss--Kuzmin heuristic accounts for only part of the observed rate. A proof would require understanding the joint distribution of continued-fraction partial quotients and denominators for integer logarithms, including correlations not captured by existing theory.

Progress summary

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

Sources & referencesView supporting material

Primary source

James M. Hyman, “Why Eight Percent of Benford Sequences Never Converge”, arXiv:2603.18243 (2026).

Solutions 0

No solutions have been posted yet.