Search

The Kadison–Ringrose cohomology conjecture

Proves bounded Hochschild cohomology of von Neumann algebras vanishes in degrees two and up, settling Kadison–Ringrose.

1 paper September 23, 2026 Explained by AI

A von Neumann algebra is a family of operators closed under sums, products and a strong kind of limit. A cocycle is a multi-input map that satisfies a fixed cancellation rule, and the easy ones come from simpler maps one level down. Kadison and Ringrose conjectured that every bounded cocycle is one of the easy ones. Degree one was settled in 1966 by Kadison and Sakai. This paper proves every higher degree, with no extra conditions on the algebra, and Lean checks it.

First step
1967
The self-coefficient cohomology question, which Sinclair and Smith trace to 1967; Kadison and Ringrose developed the higher-degree theory
Significance
Landmark
It settles the Kadison–Ringrose cohomology conjecture for all von Neumann algebras, with a Lean check of the higher degrees.
Trust
Formal proof listed
Field
Operator algebras
1 paper

The explanation

Take 2-by-2 matrices and write e11, e12, e21 for the matrices with a single 1 in that position. Fix b = e11 and set D(a) = ab − ba. Then D(e12) = e12·e11 − e11·e12 = 0 − e12 = −e12. Also D(e21) = e21 and D(e11) = 0. Check the product rule D(xy) = xD(y) + D(x)y with x = e12 and y = e21. The left side is D(e11) = 0. The right side is e12·e21 − e12·e21 = e11 − e11 = 0. They agree.

A map with that product rule is a derivation. The rule is the degree-one cocycle equation. A cocycle is a bounded map of k inputs whose Hochschild differential, a fixed alternating sum, is zero. Bounded means it stretches inputs by at most a fixed factor. Cocycles of the form D(a) = ab − ba come from one element b, one degree lower. Such cocycles are called coboundaries, or inner when k = 1. Bounded Hochschild cohomology in degree k measures cocycles that are not coboundaries. When it is zero, every cocycle equation has a bounded solution one degree down.

The algebras here are von Neumann algebras: families of operators on a Hilbert space, closed under sums, products, adjoints and pointwise limits. Kadison and Sakai showed every derivation of such an algebra into itself is inner. The Kadison–Ringrose conjecture asks the same in every degree. Earlier work settled many cases: type I and hyperfinite algebras, algebras with a Cartan subalgebra, and factors with property Γ. The obstacle was the general type II₁ case, with no extra structure to average over.

A primitive is a map one degree lower whose differential is the cocycle. The paper proves every bounded cocycle of degree at least two has a bounded primitive. It needs no separability or type restriction. The key step averages one input over a single random walk on the algebra's unitaries, the operators that preserve length. Averaging gives a harmonic map, one equal to its own average. A new Liouville theorem shows such a map must have the simple form T(x) = T(1)x. Together with the degree-one theorem, this settles the conjecture positively.

How sure is this? Lean, a computer proof checker, has a formal statement of the main theorem: every bounded cocycle of degree at least two has a bounded primitive. The degree-one theorem is outside that statement, but it is a 1966 result. So the paper proves the higher-degree theorem, as far as the Lean statement matches the paper. No specialist review is published.

What it doesn't show. This does not reprove the degree-one inner-derivation theorem; it relies on that earlier result for degree one.

Where it came from

This result comes from an unreleased OpenAI model, published in the openai/math release on 6 October 2026.

Hype check

Landmark-level claim with a Lean check of every degree from two up. Degree one rests on the 1966 Kadison–Sakai theorem. No specialist review is published.

Words used

Von Neumann algebra
A family of operators on a Hilbert space closed under sums, products, adjoints and pointwise limits.
Derivation
A linear map D with the product rule D(xy) = xD(y) + D(x)y.
Inner derivation
A derivation of the form D(a) = ab − ba for one fixed element b.
Hochschild cocycle
A map of k inputs whose Hochschild differential, a fixed alternating sum built from products, is zero.
Primitive
A map g of one fewer input whose differential dg equals the given cocycle.
Bounded cohomology H_b^k
The bounded degree-k cocycles modulo those with a bounded primitive; zero means every cocycle has one.
Type II₁
A class of von Neumann algebras with a finite trace and no minimal projections; the hardest case here.
Unitary
An operator U with U*U = UU* = 1, so it preserves lengths.

The original paper

Lean notes

All quoted lines (9)
  1. The continuous Hochschild cohomology of an operator algebra measures the obstruction to solving its bounded cocycle equations.
    Vanishing of higher bounded Hochschild cohomology Back to text
  2. For a von Neumann algebra acting on itself by multiplication, the degree-one equation is the derivation equation, and its vanishing is the inner-derivation theorem of Kadison and Sakai.
    Vanishing of higher bounded Hochschild cohomology Back to text
  3. The Kadison–Ringrose cohomology conjecture asks whether Hbk(M,M)=0H_b^k(M,M)=0 for every complex von Neumann algebra MM and every positive integer kk.
    Vanishing of higher bounded Hochschild cohomology Back to text
  4. Kadison and Ringrose proved vanishing for type I and hyperfinite von Neumann algebras.
    Vanishing of higher bounded Hochschild cohomology Back to text
  5. Christensen, Pop, Sinclair, and Smith proved vanishing for all type II1\mathrm{II}_1 factors with property Γ\Gamma, including nonseparable ones.
    Vanishing of higher bounded Hochschild cohomology Back to text
  6. We prove that every bounded Hochschild cocycle of degree at least two on a complex von Neumann algebra, with values in the algebra itself, has a bounded primitive.
    Vanishing of higher bounded Hochschild cohomology Back to text
  7. We obtain the needed module identity by averaging over a single unitary random walk.
    Vanishing of higher bounded Hochschild cohomology Back to text
  8. Averaging one cocycle input first gives only a harmonic map; the main new input is a Liouville theorem that makes such a map commute with right multiplication.
    Vanishing of higher bounded Hochschild cohomology Back to text
  9. Together with the established degree-one inner-derivation theorem, this resolves the Kadison–Ringrose cohomology conjecture positively.
    Vanishing of higher bounded Hochschild cohomology Back to text

Other sources

  1. openai/math, Lean scope note for family 295 The degree-one inner-derivation theorem is outside this selected statement.
  2. openai/math release notes (2026): model provenance and review status manuscripts and supporting proof artifacts produced by an internal OpenAI model

Browse Operator algebras Named conjecturesOpen since the 1960sLean-checked resultsLandmark results