110-Page Proof Settles a Quantum Entropy Conjecture — and Maps Exactly Where Strong Subadditivity Breaks

Subadditivity proven, strong subadditivity disproven for every entropy between von Neumann and subentropy. A definitive resolution with an operational payoff in channel capacity.

Imagine you have a set of Russian nesting dolls, and you know a rule: the weight of any pair of nested dolls is never more than the sum of their individual weights. That's subadditivity — a basic accounting constraint on how 'information content' adds up across subsystems. Now ask a harder question: does this accounting constraint also hold when you condition on a third doll sitting inside? That's strong subadditivity, and for the standard quantum entropy (von Neumann), it's been a bedrock theorem since Lieb and Ruskai proved it in 1973. This paper asks: what happens for the *other* entropies in the interpolating family that stretches from von Neumann down to subentropy? The committed claim: subadditivity holds for the entire Nichols–Wootters (NW) family of entropies, settling the open conjecture for subentropy specifically, but strong subadditivity fails for every member except von Neumann entropy. The failure is demonstrated concretely — a maximally entangled two-qubit state tensored with a mixed qubit already breaks it. This is not a marginal or exotic counterexample; it's a bread-and-butter quantum state. The architecture is pure mathematical proof, operating within the operator-algebraic and matrix-analytic framework standard in quantum information theory. No numerics, no simulation, no hardware. The central technical engine is a 'positive representation by Haar averages of compressed matrix entropies,' which yields a pinching inequality that cascades into the subadditivity result and also produces entropy-gain bounds for unital quantum channels. The Haar-average representation is the load-bearing structural innovation — it is what lets the authors handle all spectra in a common ambient dimension simultaneously. On the ladder: the von Neumann entropy's strong subadditivity (Lieb–Ruskai 1973) is the classical baseline this paper is measured against. The NW family was introduced by Nichols and Wootters as a natural interpolation, but the subadditivity conjecture for subentropy remained open. No prior work had resolved it. For strong subadditivity, the paper doesn't just show failure — it computes exact defects, identifies both-sign behavior on quantum Markov states, and determines optimal coefficients for two-qubit states in every ambient dimension (including the subentropy coefficient 1/3). This is not an incremental push past a prior bound; it is a complete resolution. Integrity is as strong as it gets for a pure-math paper: the results are proved, not simulated. The counterexamples are explicit and constructable. The equality conditions for subadditivity are tight (equality iff at least one marginal is pure). The strong-subadditivity defect's exact maximum under rank constraints is determined, not estimated. The only caveat — inherent to the genre — is the absence of independent verification, though the 110-page length with detailed proofs invites checking. The operational payoff is not an afterthought. Each NW entropy equals the classical communication capacity retained after a specified isotropic partial projective measurement, at a prescribed average input state, even when entangled block encodings are allowed. This converts a family of abstract entropy quantities into channel-capacity numbers with direct physical meaning — a bridge between mathematical structure and information-theoretic function that justifies caring about these entropies beyond curiosity. At 110 pages, this is a monograph-scale effort. The authors do not leave loose ends within their stated scope. The obvious successor experiments are extending the framework to multipartite settings beyond the tripartite counterexample, and testing whether the defect structure has consequences for quantum error correction or entanglement distillation protocols. The paper's own application section hints at operational directions but does not pursue them quantitatively beyond the capacity interpretation.