Reference

Glossary

Plain-language meanings for the terms used on this site. In the chapters, any word with a dotted underline opens its entry in a pop-up.

Terms

absolute value
The size of a number, ignoring its direction. For a complex number it is the distance from zero on the complex plane. It is 0 only when the number itself is 0. Written |z|. Wikipedia ↗
analytic continuation
A way of extending a function beyond where its original formula works. The zeta sum only makes sense when the real part of s is bigger than 1, but there is exactly one natural way to extend it to (almost) all complex numbers, and that extension is what “ζ” means. Wikipedia ↗
Chebyshev's function ψ(x)
The running total of the von Mangoldt weights Λ(n) for all n up to x. It counts primes (and prime powers) with each one weighted by ln p, and it stays close to x itself. Wikipedia ↗
completions of a number field
For each prime there is a different way of measuring how close two numbers are, plus the ordinary way. Each gives a filled-in version of the number system. Working with all of them at once (the “adeles”) is the standard modern way to treat every prime on an equal footing. Wikipedia ↗
complex number
A number with two parts, written a + bi, where i is a new number whose square is −1. Think of it as a point on a map: a is how far east, b is how far north. Ordinary numbers are the ones with b = 0. Wikipedia ↗
complex plane
The map on which complex numbers live: real part along the horizontal axis, imaginary part along the vertical axis. Wikipedia ↗
convexity bound
A standard, already-proved estimate for how large an L-function can get inside the critical strip. Weaker than what RH would give, but known unconditionally.
convolution
Blurring: replace each value of a function with a weighted average of nearby values, with the weights given by a bump. Wikipedia ↗
critical line
The vertical line down the middle of the critical strip: complex numbers with real part exactly ½. The Riemann Hypothesis says every interesting zero of zeta is on this line. Wikipedia ↗
critical strip
The vertical band of complex numbers whose real part is between 0 and 1. All the interesting zeros of zeta are inside it. Wikipedia ↗
Dirichlet character
A repeating pattern of numbers (such as +1, −1, +1, −1…) attached to the whole numbers, which respects multiplication. Putting one into the zeta sum gives a Dirichlet L-function, used to study primes in patterns like “one more than a multiple of 4”. Wikipedia ↗
eigenfunction
A function that an operator sends to a constant multiple of itself. The constant is the eigenvalue. Wikipedia ↗
eigenvalue
Some inputs come out of an operator unchanged except for being multiplied by a number. That number is an eigenvalue. A pure tone passed through a blur comes out as the same tone, just quieter; “how much quieter” is the eigenvalue. Wikipedia ↗
entire function
A function of a complex number that is smooth everywhere on the complex plane, with no poles or other bad points. Wikipedia ↗
Euler product
Euler's discovery that the zeta sum over all whole numbers equals a product with one factor for each prime. It works because every whole number factors into primes in exactly one way. Wikipedia ↗
Euler's constant
A fixed number, about 0.5772, that turns up whenever sums like 1 + 1/2 + 1/3 + … are compared with logarithms. Usually written γ (not to be confused with the γ used for the height of a zero). Wikipedia ↗
exact sequence
A compact way of recording “this small space sits inside that big one, and here is what is left when you divide it out”. Wikipedia ↗
explicit formula
An exact equation that writes the prime staircase as a smooth main term minus one wave for each zero of zeta. It is the precise link between primes and zeros. Wikipedia ↗
formalization
A proof written out in full detail in a language like Lean, so a computer can verify it. Much longer than a paper proof, but nothing is left to trust. Wikipedia ↗
Fourier mode
One of the pure oscillations that fit a whole number of times around a circle. Every reasonable function on the circle is a sum of them (its Fourier series). Wikipedia ↗
Fourier transform
The mathematical version of splitting a sound into its pure tones: it rewrites a function as a combination of simple waves and tells you how much of each frequency is present. Wikipedia ↗
functional equation
A formula relating ζ(s) to ζ(1 − s). Its practical meaning: the zeros of zeta are arranged symmetrically, as mirror images across the critical line. Wikipedia ↗
generalized Riemann Hypothesis (GRH)
The same statement as RH for a whole family of zeta-like functions (L-functions). It controls how primes are spread among patterns such as “primes ending in 3”. Wikipedia ↗
Hecke character and Hecke L-function
The version of Dirichlet characters and their L-functions that works over any number field, not just the ordinary whole numbers. Masullo's paper claims GRH for all of them. Wikipedia ↗
Hilbert space
A space of functions with a notion of length and angle, like ordinary 3D space but with infinitely many directions. Having lengths is what makes sums of infinitely many terms, and traces, behave. Wikipedia ↗
integral (∫)
A running total of a quantity that changes continuously: the area under its curve. ∫ from a to b of f(t) dt adds up f(t) for every t between a and b. Wikipedia ↗
integral kernel
The function K(x, y) in an operator of the form f ↦ ∫ K(x, y) f(y) dy. It plays the role of the entries of a matrix, and the operator's trace is the integral of K(x, x). Wikipedia ↗
kernel check
Lean's final, independent verification of a finished proof by its small trusted core program. Until this passes, a Lean proof has not actually been verified.
L-function
A relative of the zeta function: the same kind of sum, but with each term multiplied by a repeating pattern of coefficients. Each has its own zeros and its own version of the Riemann Hypothesis. Wikipedia ↗
Landau's theorem
A classical result (1905) with this consequence: if a prime-counting error stays below a ceiling of a certain size, on one side only, then zeta has no zeros to the right of the matching vertical line. You do not need to control both sides.
Lean
A programming language for writing mathematical proofs so that a computer can check every step. If the checker accepts a proof, each step follows from the definitions and axioms. Wikipedia ↗
linear operator
A rule that turns one function into another and respects adding and scaling. Blurring a photo is an example: blur two images and add them, or add them and then blur, and you get the same result. Wikipedia ↗
logarithmic integral li(x)
The running total of 1 / ln t for all t from 2 up to x. Since about one number in ln t is prime near t, this total is the natural smooth estimate for how many primes there are up to x. Wikipedia ↗
logarithmic scale
An axis where each step multiplies instead of adding: 1, 10, 100, 1000 are evenly spaced. It lets one picture show both small and very large numbers. Wikipedia ↗
Mathlib
The large shared library of already-proved mathematics for Lean. New Lean proofs build on it instead of starting from scratch.
Mellin transform
A way of breaking a function on the positive numbers into pure “power” pieces t^s, the same way a sound is broken into pure tones. It is the Fourier transform done on a logarithmic scale. Its key property: stretching the function just multiplies its transform by a simple factor. Wikipedia ↗
multiplicity
How many times a zero counts. The function (x − 3)² is zero at 3 “twice over”, so that zero has multiplicity 2. Every zero of zeta ever computed has multiplicity 1, but that is not proven in general. Wikipedia ↗
natural logarithm (ln)
ln x answers the question “e to what power gives x?”, where e ≈ 2.718. For a feel of its size: ln x is about 2.3 times the number of digits of x, so it grows very slowly. ln 10 ≈ 2.3, ln 1000 ≈ 6.9, ln 1,000,000 ≈ 13.8. Wikipedia ↗
norm
A way of assigning a size to a function. Changing the norm changes which functions count as small, and therefore which infinite sums settle down. Wikipedia ↗
number field
A number system made by adding a few extra numbers (like √2 or i) to the fractions. Each has its own “primes” and its own zeta-like functions. Wikipedia ↗
one-sided bound
A limit on a quantity from one direction only: “never above this ceiling”, with nothing said about how low it can go (or the reverse). Wikipedia ↗
Poisson summation
An identity: adding up a function's values at evenly spaced points gives the same answer as adding up its Fourier transform at evenly spaced frequencies. It is why a trace can be computed two ways. Wikipedia ↗
pole
An input where a function blows up to infinity, like 1/x at x = 0. Zeta has exactly one, at s = 1. Wikipedia ↗
prime
A whole number bigger than 1 that cannot be divided evenly by anything except 1 and itself: 2, 3, 5, 7, 11, 13… Wikipedia ↗
prime number theorem
Proved in 1896: the number of primes up to x is approximately x / ln x, and the approximation gets better (as a percentage) the larger x is. Wikipedia ↗
prime power
A prime multiplied by itself some number of times: 2, 4, 8, 16… or 3, 9, 27… A prime itself counts as its own first power. Wikipedia ↗
product sign (Π)
Π means “multiply together”, in the same way Σ means “add up”. Π over primes p means one factor for each prime.
quotient space
What you get by declaring certain things equal and then forgetting the difference. Clock arithmetic is a quotient: 13 o'clock and 1 o'clock are declared the same. Here, two functions are declared the same when they differ by a θ-function. Wikipedia ↗
real part and imaginary part
The two coordinates of a complex number a + bi: a is the real part (left–right position), b is the imaginary part (up–down position).
Riemann Hypothesis (RH)
The conjecture, made by Riemann in 1859, that every zero of zeta inside the critical strip has real part exactly ½. Equivalent to saying the primes are spread as evenly as they possibly could be. Wikipedia ↗
ringing
The wiggle and overshoot you get next to a sharp jump when you try to build it out of smooth waves. Also called the Gibbs phenomenon; it is why a square wave built from sine waves has little horns at its corners. Wikipedia ↗
sorry (Lean)
A placeholder in Lean meaning “proof missing here”. Lean accepts the file but marks everything that depends on it as unproved.
spectral side and geometric side
The two halves of a trace formula. Spectral: the sum of eigenvalues, one per frequency (or per zero). Geometric: a sum over closed loops (or over primes). The formula says they are equal.
sum sign (Σ)
Σ means “add up”. The small print underneath says which things to add: Σ over n ≤ x means add one term for every whole number n up to x. Wikipedia ↗
Tate's thesis
John Tate's 1950 doctoral thesis, which re-derived the main properties of zeta and L-functions using Fourier analysis over all primes at once. It is the starting point for Masullo's approach. Wikipedia ↗
test function (bump)
A smooth hill-shaped function that is exactly zero outside a limited range. It is a probe you get to choose: different bumps make a formula reveal different things. Wikipedia ↗
topology (of a function space)
A precise notion of when two functions count as “close”. Different notions of closeness give genuinely different spaces, and a statement can be true for one and false for another. Wikipedia ↗
trace
One number that summarizes an operator: the sum of all its eigenvalues. The same number can also be computed a completely different way, by adding up how strongly the operator connects each point to itself. Wikipedia ↗
trace formula
An equation obtained by computing one trace in two ways: once as a sum over frequencies (the “spectral side”), once as a sum over ways of returning to where you started (the “geometric side”). Wikipedia ↗
trivial zeros
The zeros of zeta at −2, −4, −6, … They are easy to find and explain, and have no effect on the prime-counting question. “Zeros of zeta” usually means the other ones. Wikipedia ↗
von Mangoldt function Λ(n)
A weight given to each whole number n. If n is a power of a prime p (p, p², p³…), the weight is ln p. Every other number gets weight 0. It is the bookkeeping device that makes prime-counting formulas come out clean. Wikipedia ↗
zero (of a function)
An input where a function's output is 0. The zeros of zeta are the complex numbers s where ζ(s) = 0. Wikipedia ↗
zeta function ζ(s)
The function ζ(s) = 1 + 1/2ˢ + 1/3ˢ + 1/4ˢ + … extended to complex numbers s. It encodes all the primes at once through Euler's product, and the places where it equals zero control how evenly primes are spread. Wikipedia ↗
θ-function (theta function)
A function made by adding up the values of another function along all the multiples of a point: Σ f(nx). The sum over whole numbers is what links it to the zeta function. Wikipedia ↗

People

Alain Connes
French mathematician (born 1947), Fields Medal 1982. In the 1990s he proposed approaching the Riemann Hypothesis through a trace formula on a space built from all primes; Masullo's paper starts from that picture. Wikipedia ↗
Bernhard Riemann
German mathematician (1826–1866). His single 1859 paper on primes extended zeta to complex numbers, related primes to its zeros, and stated the hypothesis that carries his name. Wikipedia ↗
Carl Friedrich Gauss
German mathematician (1777–1855). As a teenager he guessed from tables of primes that about one number in ln x is prime near x, which is the starting point of this whole subject. Wikipedia ↗
Edmund Landau
German mathematician (1877–1938) who systematized the theory of prime numbers. One of his theorems is why a bound on one side is enough. Wikipedia ↗
Hans von Mangoldt
German mathematician (1854–1925) who supplied rigorous proofs of formulas Riemann had only sketched, including the explicit formula. Wikipedia ↗
Hjalmar Mellin
Finnish mathematician (1854–1933), for whom the Mellin transform is named. Wikipedia ↗
John Tate
American mathematician (1925–2019). His 1950 thesis rebuilt the theory of zeta and L-functions using Fourier analysis over all primes at once. Wikipedia ↗
Leonhard Euler
Swiss mathematician (1707–1783). He discovered that the zeta sum equals a product over the primes, the first link between primes and a smooth function. Wikipedia ↗
Pafnuty Chebyshev
Russian mathematician (1821–1894). He introduced the weighted prime counts used throughout these chapters and proved the first solid bounds on how many primes there are. Wikipedia ↗
Peter Borwein
Canadian mathematician (1953–2020). He found a fast, simple method for computing the zeta function, which the labs on this site use. Wikipedia ↗
Ralf Meyer
German mathematician who, following Connes, studied the function spaces behind this approach in the 2000s.
Siméon Denis Poisson
French mathematician (1781–1840), for whom the Poisson summation formula is named. Wikipedia ↗