The exceptional set for the distribution of primes between consecutive powers (Q2460679)

From the text: A well known conjecture about the distribution of primes asserts that between two consecutive squares there is always at least one prime number. The proof of this conjecture is quite out of reach at present, even under the assumption of the Riemann Hypothesis. This paper is concerned with the distribution of prime numbers between two consecutive powers of integers, as a natural generalization of the afore-mentioned conjecture. The well known result of \textit{M. N. Huxley} [Invent. Math. 15, 164--170 (1972; Zbl 0241.10026)] about the distribution of primes in short intervals implies that, for \(n\to\infty\), all intervals \([n^\alpha, (n+1)^\alpha]\) contain the expected numbers of prime for \(\alpha > \tfrac {12}{5}\) and ``almost all'' contain the expected number of primes for \(\alpha > \tfrac 65\). By ``almost all'' it is meant that the number of integers \(X\leq n\leq 2X\) for which the interval \([n^\alpha, (n+1)^\alpha]\) does not contain the expected number of primes is \(o(X)\). The author's first result is an unconditional estimate for the measure of the exceptional set for the distribution of primes between two consecutive powers of integers. Theorem 1. Let \(\varepsilon> 0\). Then for every \([n^\alpha, (n+1)^\alpha]\subset [N,2N]\) with \(O((N^{1/\alpha})^{\eta(\alpha)+\varepsilon})\) exceptions we have the expected number of primes, where \[ \eta(\alpha)=\begin{cases}\\ \frac 85 - \frac \alpha 2,\qquad & \frac 65 <\alpha\leq \frac 65 +c,\\ \frac 52 -\alpha, & \frac{27}{16}<\alpha\leq\frac {53}{26}, \\ \frac{72-9\alpha-8\alpha^2}{3(\alpha+12)}, & \frac{53}{26}\leq\alpha < \frac{12}5 \end{cases} \] and \(c\) is a suitable positive constant. For the sake of simplicity we consider the function \(\eta(\alpha)\) only for the extremes and more interesting values of the parameter \(\alpha\). Arguing in the same way we can obtain the explicit values of the function \(\eta(\alpha)\) for every \(\alpha\). A corollary of this theorem is Theorem 1 of the author's paper [Arch. Math. 75, No. 1, 29--34 (2000; Zbl 1047.11087)], which states that for every \([n^2, (n+1)^2] \subset [N, 2N]\) with \(O\left(N^{1/4+\varepsilon}\right)\) exceptions we have the expected number of primes. Assuming some usual hypotheses the author obtains strong results about the regularity of distribution of primes and then a better estimate for the number of exceptions for the distribution of primes between two consecutive powers of integers. First he assumes the Lindelöf hypothesis, which states that, for every \(\eta> 0\), the Riemann zeta-function satisfies \[ \zeta(\sigma + it)\ll t^\eta\quad (\sigma\geq \tfrac 12, t\geq 2). \] Assuming this hypothesis he proves that all intervals \([n^\alpha, (n+1)^\alpha]\) contain the expected number of primes for \(\alpha > 2\) and almost all contain the expected number of primes for \(\alpha > 1\), see the author [Rocky Mt. J. Math. 39, No. 2, 413--421 (2009; Zbl 1193.11085)]. His second result is a conditional estimate for the measure of the exceptional set for the distribution of primes between two consecutive powers of integers under the assumption of the Lindelöf hypothesis. Theorem 2. Assume the Lindelöf hypothesis, let \(\varepsilon > 0\) and \(1 < \alpha < 2\). Then for every \([n^\alpha, (n+1)^\alpha] \subset [N, 2N]\) with \(O((N^{1/\alpha})^{2-\alpha+\varepsilon})\) exceptions we have the expected number of primes. A corollary of this theorem is Theorem 2 of the author [2009, op. cit.], which states that almost all intervals \([n^\alpha, (n+1)^\alpha]\) contain the expected number of primes for \(\alpha > 1\). Assuming the Riemann Hypothesis his third result is as follows. Theorem 3. Assume the Riemann Hypothesis, let \(1<\alpha\leq 2\) and \(g(x)\to\infty\) arbitrarily slowly. Then for every \([n^\alpha, (n+1)^\alpha] \subset [N, 2N]\) with \(O((N^{1/\alpha})^{2-\alpha}\log^2N g(N))\) exceptions we have the expected number of primes. A corollary of this theorem is Theorem 2 of the author [2000, op. cit.], which states that for every \([n^\alpha, (n+1)^\alpha] \subset [N, 2N]\) with \(O(g(N)\log^2 N)\) exceptions we have the expected number of primes. Finally the relation with results of Heath-Brown on differences between consecutive primes are discussed.