Arf invariant
In mathematics, the Arf invariant is a value that describes a special kind of quadratic form over a field of characteristic 2. The Turkish mathematician Cahit Arf defined it when he began studying quadratic forms over these fields. In characteristic 2, it plays a role similar to the one the discriminant for quadratic forms plays in other characteristics. Arf used this invariant, along with other tools, to sort quadratic forms into groups.
For the field F2, which has just two elements, the Arf invariant is the value that appears most often among the form's values. Two such quadratic forms over F2 are the same in structure if and only if they have the same dimension and the same Arf invariant. Leonard Dickson essentially knew this fact, even for any finite field of characteristic 2. Arf later proved it for any perfect field.
The Arf invariant is especially useful in geometric topology. It is mainly used there to define a value for shapes called (4k + 2)-dimensional manifolds. These include singly even-dimensional manifolds such as surfaces, 6-dimensional shapes, and 10-dimensional shapes. These shapes need an extra feature called a framing. This idea leads to the Arf–Kervaire invariant and the Arf invariant of a knot.
The Arf invariant is similar to the signature of a manifold, which is defined for 4k-dimensional shapes, the doubly even-dimensional ones. This pattern repeating every four steps matches the 4-fold repeating pattern of L-theory. The Arf invariant can also be defined, in a more general way, for certain 2k-dimensional manifolds.
Definitions
The Arf invariant is defined for a quadratic form q over a field K of characteristic 2. The form must be nonsingular. This means its partner form is nondegenerate. Because K has characteristic 2, that partner form is alternating. So a nonsingular quadratic form in characteristic 2 must have an even number of dimensions.
Every nonsingular form with two dimensions can be written as ax² + xy + by². Its Arf invariant is the product ab. Equivalent forms give products that differ only by values of the form u² + u. So the Arf invariant does not change when a form is swapped for an equivalent one.
Every nonsingular form can be split into a sum of these two-dimensional forms. Arf showed this splitting, though Dickson had noticed it earlier for finite fields. The Arf invariant of the whole form is the sum of the invariants of its parts. This means it is additive.
Artin–Schreier theory connects these invariants to Galois cohomology. Each nonzero Arf invariant matches a separable quadratic extension field of K. So the Arf invariant is either zero or describes such a field. This is similar to the discriminant, which does the same job for fields that are not of characteristic 2. The discriminant is linked to these ideas through Kummer theory.
Arf's main results
If the field K is perfect, then every nonsingular quadratic form over K is determined, up to equivalence, by its dimension and its Arf invariant. This is true over the field F2, where the Arf invariant is either 0 or 1.
If the field K is not perfect, meaning K is different from its subfield K2 of squares, then the Clifford algebra is another important tool for describing the form. A corrected version of Arf's statement says that if the degree [K: K2] is at most 2, then every quadratic form over K is completely described by its dimension, its Arf invariant, and its Clifford algebra. Examples of such fields include function fields (or power series fields) of one variable over perfect base fields.
Quadratic forms over F2
Over F2, the Arf invariant is 0 if the quadratic form is made from copies of xy joined together. It is 1 if the form is x2+xy+y2 joined with some copies of xy.
William Browder has called the Arf invariant the democratic invariant, because it is the value that a quadratic form takes most often. Another way to think about it is that the Arf invariant is 0 when the vector space over F2 has a special part that is half its size, called a totally isotropic subspace. On this part, the form is always 0. In other words, the Arf invariant is 0 when the form's isotropy index is half of its dimension.
In topology, the study of shapes, the Arf invariant is a way to describe a manifold. It uses the intersection form and a quadratic refinement. Mathematicians split the manifold's homology into small pieces of two kinds and count how many of one kind there are. The result is a single number, either 0 or 1.
The idea shows up in several examples. For a surface of a certain genus, the Arf invariant shows whether the surface can be the edge of a 3-dimensional shape that keeps the given framing. For a knot, it is found from a Seifert surface. It does not change when a different Seifert surface is chosen, which makes it a knot invariant. It also adds up under connected sum and is zero for slice knots. In surgery theory, the Arf invariant is called the Kervaire invariant. It helps mathematicians sort shapes such as exotic spheres.
The Arf invariant in physics
The simple model of a Kitaev chain can be described by a topological quantum field theory when it is in a special state. This is a way of describing the chain's behavior at low energy using a effective field theory. In this state, a chain on a short line has special particles at its ends, and the Arf invariant describes this behavior.
In this case, a certain value of the Arf invariant matches a certain value of a mathematical tool called the mod 2 Atiyah-Singer index of the Dirac operator.

Quiz time!
4 questions about Arf invariant
Who defined the Arf invariant when he began studying quadratic forms over these fields?
Keep exploring
This article is a child-friendly adaptation of the Wikipedia article on Arf invariant, available under CC BY-SA 4.0.
Images from Wikimedia Commons. Tap any image to view credits and license.