How statement and proof provenance work
The first chip identifies the source of the statement or construction; the second identifies the source of its local proof or verification.
- Literature-sourced: the exact statement appears in a cited source; only wording and notation differ.
- AI-adapted: a semantically identical restatement of literature-sourced material, modulo indexing, notation, and boundary cases adopted by the library.
- AI-generated: a genuinely novel statement formulated by AI, with no source for the claim itself.
These labels describe origin, not correctness: citations and verification chips remain separate evidence.
Boundary energy bounds thermal mutual information
Statement
Assume the Axiom of Choice. Let be nonzero finite-dimensional complex inner-product spaces. Let act on , let act on , and let act on . Set
For , define and the finite Gibbs density . Then
where uses dimensionless entropy in nats and is the operator norm.
For the optional lattice-family conclusion, fix . For each member of a specified family, take a finite site set , a partition , and nonzero finite-dimensional complex inner-product spaces for . Use the lexicographic order on in , with the empty tensor product defined as , and identify with in the displayed factor order. Let be a finite set of interaction labels with nonempty supports and self-adjoint terms on . For containing , write for the extension of to by the identity on every factor in . The crossing-label set is
Let and . Set , , and . Under the stated identification these give the displayed Hamiltonian decomposition. For each , assume and define by the Gibbs formula above. If one specified satisfies for every family member, the bound gives memberwise. No finite-range condition or additional geometric scaling assumption is used. When and each interaction term are measured in joules, has units J and has units J, so the bound is dimensionless.
Statement references
Facts & Assumptions
Given: The spaces and Hermitian matrices in the Statement, , natural logarithms, and a fixed with units J/K when physical units are assigned. Put .
The Axiom of Choice: every family of nonempty sets has a choice function. This is inherited through the Gibbs variational supplier in [F1] and its [F2] relative-entropy-domain dependency Quantum spectral entropy and relative-entropy domains; the finite matrix calculation in this item makes no infinite-dimensional selection.
Finite Gibbs free-energy and entropy optimization: for a finite Hermitian Hamiltonian, its Gibbs density minimizes over all density matrices, including rank-deficient competitors, with and . Its relative-entropy-domain dependency, recorded as [F2] in that supplier, carries Choice through the declared supplier chain; its own [A1] records energy and units.
Partial traces of a finite bipartite state are states: both marginals of a finite bipartite density operator are density operators.
Finite partial trace, Schmidt decomposition and finite no-signalling: for every finite bipartite endomorphism and local endomorphism , , with the corresponding identity for tracing the first factor.
Finite matrix spectral and state calculus: finite Hermitian matrices have orthonormal eigenbases with real eigenvalues; positive matrices have nonnegative spectral weights, and the finite trace is cyclic.
Finite mixing, product entropy, and bipartite subadditivity: for finite-dimensional density matrices, the physical entropy of a tensor product is the sum of the physical entropies of its factors, including rank-deficient factors. Its declared supplier chain carries the same Choice hypothesis in [A1].
Proof
By [F5], has finitely many real eigenvalues, so is positive definite and . Thus is a finite Gibbs density to which [F1] applies. Its marginals and are density operators by [F2].
Set . Tensor products of orthonormal eigenbases show that is positive semidefinite with eigenvalues , whose sum is ; hence is a density operator. By [F6], . Since and , division by gives . Therefore [F4] gives .
Define for finite density matrices. Since , [F1] applies to this functional and yields . Rearranging and using step 2.1 gives .
The coordinate formula in [F3] gives and , since each factor has trace one. The defining trace-pairing identities in [F3], with cyclicity from [F5], then show that the expectations of in and agree, and likewise for . Thus both local terms cancel in the trace on the right of step 3.1, leaving .
Diagonalize by [F5]. For any density matrix , the numbers are nonnegative (each is a sum of nonnegative diagonal expectations of on an orthonormal basis of ) and sum to one. Thus . Applying this absolute bound to and gives . Combine this with steps 3.1 and 4.1 and multiply by to obtain the claimed inequality. Substituting the additional family bound yields its stated conditional area scaling.
Depends on
The Axiom of Choice
Dimensionless entropy, mutual and conditional information
Quantum spectral entropy and relative-entropy domains
Partial trace over a finite tensor factor
Partial traces of a finite bipartite state are states
Finite matrix spectral and state calculus
Finite mixing, product entropy, and bipartite subadditivity
Finite partial trace, Schmidt decomposition and finite no-signalling
Finite Gibbs free-energy and entropy optimization
Dependency tree · 2 levels
Sources
- Michael M. Wolf, Frank Verstraete, Matthew B. Hastings, and J. Ignacio Cirac, Area laws in quantum systems: mutual information and correlations, arXiv:0704.3906v2, ‘Area laws for classical and quantum systems,’ Eq. (3)
- Pieter Naaijkens, Quantum Spin Systems on Infinite Lattices, §3.4, finite-dimensional entropy and Gibbs free-energy identity
- John Watrous, The Theory of Quantum Information, Chapter 5, §§5.2.1–5.2.3, author-hosted manuscript