A zero-density wall, operated
The Wall at Three Quarters
Move a line through the zeta critical strip and watch three published zero-density exponents compete. At sigma = 3/4, the live calculation changes the classical exponent from 3/5 to 5/9. A second instrument carries the uniform density coefficient into strict short-interval thresholds, then runs an exact finite sieve without pretending that finite data proves an asymptotic theorem.
The curves do not locate a zeta zero. They compare proven asymptotic upper-bound exponents for how many zeros could sit on or to the right of your line.
Layer one: the critical strip
Drag the wall
The lowest curve is the strongest exponent-level comparison at that value of sigma. The teal improvement exists only between the two live crossings.
These are asymptotic exponent comparisons. The browser reports E times log10(T), never a certified finite cap T to the E with coefficient one. Every theorem exponent carries an o(1) term.
A small exponent reaches the primes
The sharper teal curve is not replaced here. The next instrument uses the paper's simpler uniform envelope, A(1-sigma), because two classical bridges from a uniform coefficient A produce live threshold exponents for short intervals.
Layer two: density to intervals
Carry the coefficient across
Choose A, choose a finite x and theta, then compare the strict asymptotic gate with exact local prime counts.
Loading theorem scope.
The sieve is an exact finite computation and an illustration only. It does not verify an asymptotic threshold, identify an exceptional interval, or supply the theorem's unknown sufficiently-large-x cutoff.
The check
The page recomputes the load-bearing identities below in exact rational arithmetic, and then checks each solved crossing against the very floating-point functions that draw the curves. The rational pass and the plotted pass are written separately on purpose: agreement between them is the check.
Conventions and free choices
- N(sigma,T) counts zeta zeros rho with real part at least sigma and absolute imaginary part at most T, including multiplicity.
- All displayed density values are exponents with an o(1) term. No implicit constant is assigned a numerical value.
- The finite sieve counts primes p with x < p ≤ x + floor(xtheta). Its normalization is count divided by y/log(x), with natural logarithm.
- The five x presets, ten theta presets, thirteen neighboring intervals, plot scales, colours, and current landing date are editorial free choices. The supplied specification contained no landing date, so this page uses 2026-07-29.
- The direct primality cross-check covers the first min(y, 512) integers of the chosen interval. That overlap length is an implementation choice.
Qualifications and approximations
- The o(1) terms, implicit constants, and dependence on each fixed positive epsilon are not explicit here.
- The every-interval corollary also assumes y ≤ x0.99, sufficiently large x, and a fixed epsilon greater than zero. The endpoint is excluded.
- The almost-all corollary permits a quantified exceptional set. It does not say every interval works, and its endpoint is excluded.
- The short-interval derivation also uses a Vinogradov-Korobov zero-free region with an unspecified suitable constant c greater than zero. This page does not choose c.
- Floating-point decimals and canvas coordinates are presentational approximations. Exact rational identities are retained and printed below.
Still open
- The Riemann Hypothesis remains unproved. This result does not show that any zero lies on or off the critical line.
- The density hypothesis with coefficient A equal to 2 remains unproved.
- The corollaries do not include their threshold endpoints, do not give an effective smallest x, and do not establish the same statements below those thresholds.
Live finite cross-check
Loading the segmented and direct checks.
The offline verifier extracts the page's shipped functions, sweeps every reachable control value against independently written references, and confirms that a deliberate mutation is detected.
Primary sources, checked 2026-07-29
- Larry Guth and James Maynard, New large value estimates for Dirichlet polynomials, Annals of Mathematics 203 (2026), no. 2, 623-675, DOI 10.4007/annals.2026.203.2.6. Peer-reviewed. The journal records receipt on 8 July 2024, acceptance on 16 April 2025, and online publication on 1 March 2026. The arXiv record 2405.20552 records v1 at 00:38:45 UTC on 31 May 2024.
- A. E. Ingham, On the estimation of N(sigma,T), The Quarterly Journal of Mathematics os-11 (1940), no. 1, 201-202, DOI 10.1093/qmath/os-11.1.201. Peer-reviewed, published 1 January 1940.
- M. N. Huxley, On the difference between consecutive primes, Inventiones Mathematicae 15 (1971/72), 164-170, DOI 10.1007/BF01418933. Peer-reviewed.
- Ayla Gafni and Terence Tao, On the number of exceptional intervals to the prime number theorem in short intervals, Essential Number Theory 5 (2026), 221-241, DOI 10.2140/ent.2026.5.221, arXiv:2505.24017. Peer-reviewed journal article. Its introduction independently restates the strict 17/30 and 2/15 consequences.