Cramér model

Cramér model
Name	Cramér model
Caption	Heuristic model for prime distribution
Introduced	1936
Inventor	Harald Cramér
Field	Prime number theorem, Analytic number theory

Cramér model The Cramér model is a probabilistic heuristic introduced by Harald Cramér to describe the distribution of prime numbers and predict patterns such as prime gaps, formulated in the context of developments around the Prime Number Theorem, Riemann hypothesis, and early 20th-century probability theory. The model influenced subsequent work by G. H. Hardy, John Littlewood, Atle Selberg, and Paul Erdős, and it has been compared with numerical data from computations by Daniel Shanks, Thomas R. Nicely, and projects at Mathematical Tables and Aids to Computation.

Introduction

Cramér proposed a random model driven by ideas from Probability theory, aiming to emulate correlations in the sequence of prime numbers observed in computations by V. A. Lebesgue and theoretical results by Jacques Hadamard and Charles-Jean de La Vallée Poussin. The heuristic connects to conjectures influenced by the Riemann zeta function, the Riemann hypothesis, the Prime Number Theorem, and results by Bernhard Riemann, G. H. Hardy, and John E. Littlewood, while informing later probabilistic frameworks used by Erdős and Pál Turán.

Statement of the model

In Cramér's formulation each integer n ≥ 2 is assigned as "prime" with probability 1 / log n, inspired by the asymptotic density from the Prime Number Theorem proved by Jacques Hadamard and Charles-Jean de La Vallée Poussin, and refined by later explicit estimates of Rosser and Schoenfeld. The model treats these events as independent except for trivial parity constraints, contrasting with arithmetic constraints studied by Dirichlet in the context of Dirichlet characters and by Émile Borel in measure-theoretic probability. Cramér's original papers situate the model among contemporaneous probabilistic ideas developed by Harald Bohr and Andrey Kolmogorov.

Probabilistic predictions and prime gaps

Using independence and the density 1 / log n, Cramér predicted that the maximal gap G(x) between primes below x is typically O((log x)^2), a quantitative claim that framed discussions with results by Westzynthius, Paul Erdős, and Rankin, and contrasts with bounds derived under assumptions like the Riemann hypothesis and conjectures by Goldston and Yıldırım. The model yields probabilistic estimates for occurrences of twin prime-type configurations reminiscent of conjectures by Hardy and Littlewood, it suggests statistical distributions for gaps analogous to limit laws in Paul Lévy theory, and it implies predictions about k-tuples that motivated computational verification by Tomás Oliveira e Silva and theoretical inquiries by Ben Green and Terry Tao.

Evidence, numerical tests, and limitations

Extensive computations by Tomás Oliveira e Silva, Thomas R. Nicely, and teams at Mathematics of Computation reveal both agreement and significant deviations from Cramér's predictions, paralleling discrepancies noted by Maier in the context of primes in short intervals and by Granville who proposed modifications after studying arithmetic constraints linked to sieve theory developed by Brun and Atle Selberg. The independence assumption is contradicted by phenomena explained via Montgomery's pair correlation conjecture and links to zeros of the Riemann zeta function studied by Hugh Montgomery and Andrew Odlyzko, while large gap constructions by Rankin and improvements by Pintz show limitations of naive probabilistic independence.

Variants and refinements

Refinements include the Cramér–Granville adjustment proposed by Andrew Granville, combinatorial models influenced by Brun sieve‑type considerations from Viggo Brun, and random models incorporating correlations derived from the pair correlation framework of Hugh Montgomery and conjectures by Keating and Snaith about value distribution of L-functions. Other approaches draw on work by Granville and Martin and probabilistic sieve models developed in collaboration with James Maynard and Goldston, Pintz, Yıldırım techniques, connecting to breakthroughs by Yitang Zhang and improvements by James Maynard and Polymath collaborations.

Applications and influence in analytic number theory

Cramér's heuristic has shaped conjectures and motivated proofs in prime number theory, inspiring research on maximal gaps, distribution in short intervals, and k-tuple conjectures connected to Hardy–Littlewood conjecture. The model influenced empirical projects by Richard Brent and John Pollard and theoretical advances by Goldston, Pintz, Yıldırım, and Maynard in small gap results, while informing probabilistic perspectives used in work by Terence Tao and in interdisciplinary studies linking random matrix theory as developed by Freeman Dyson and Eugene Wigner to zero statistics of the Riemann zeta function and conjectural prime behavior. Its legacy persists in framing open problems about primes that engage researchers across institutions such as Princeton University, Cambridge University, Massachusetts Institute of Technology, and Institut des Hautes Études Scientifiques.

Category:Analytic number theory