Isomorphism of Bernoulli shifts

A Bernoulli shift assigns independent random labels to the elements of a group. The group moves the labels by translation. When do two choices of label distribution give the same measurable dynamical system?

Why it matters and what counts as progress

Why it matters: For many groups, entropy measures how much randomness a shift contains. Examples that erase distinctions between base distributions test the limits of that classification.

What counts as progress: An isomorphism must preserve probability, commute with the group action and have a measurable inverse on invariant sets of full measure. Results must specify the group and the allowed base distributions.

How the problem developed

  1. 21 May 2018 · Equal-entropy isomorphism theorem [4]

    Brandon Seward Proved that Bernoulli shifts with equal-entropy standard bases are isomorphic over every countably infinite group.

  2. 12 Nov 2026 · All-base construction [1][2]

    Peter Burton with Claude Registered a countable group for which all nontrivial standard bases give measurably isomorphic Bernoulli shifts.

AI contribution

A countable group all of whose nontrivial Bernoulli shifts are isomorphic

There is a countable group G=NG = N × (Z/2\mathbb{Z}/2 ≀ Z\mathbb{Z}) whose Bernoulli shifts are measurably isomorphic for every pair of standard probability bases that are not point masses. The isomorphism is equivariant and measure preserving on invariant sets of full measure; base entropies may differ or be infinite.

AI involvementAI-led
ImportanceNotable
Technical verificationFormal (Lean 4)
Human understandingFew
Reported

Sources

  1. Burton’s Bernoulli-shift construction
  2. Build, axiom audit and NanoDa checking record
  3. Model contributions and exact theorem scope
  4. Brandon Seward, Bernoulli shifts with bases of equal entropy are isomorphic
Publication history