Mathematicians Close the Book on Nine-Party Entanglement by Building AME States in Every Composite Dimension

Explicit constructions of absolutely maximally entangled states in dimensions 2q settle a decade-old classification for 9 parties and open a systematic route for higher party counts.

Imagine you're trying to build a team of nine jugglers who can toss any four of their balls into a shared pool and have that pool look perfectly random — no matter which four you pick. That's the structural constraint of an absolutely maximally entangled (AME) state: every possible half-or-fewer subset of parties must be maximally mixed. For nine parties in local dimension d, the existence question has been open for decades. This paper closes it. The core claim: for every integer d ≥ 3, an AME(9,d) state exists. Period. The proof is constructive — not a counting argument or a probabilistic existence proof. Pandya builds explicit state families for local dimensions of the form 2q, where q is an odd prime power. The mechanism is a qubit–qudit coupling: pair each party's two-level system with a q-level system, use the q-level part to suppress interference between branches, and then only require that the qubit part has maximally mixed reductions on average rather than individually. That relaxation is the key trick. The mathematical architecture belongs to the algebraic combinatorics family — specifically, the interplay between orthogonal Latin squares, Reed–Solomon-type codes, and stabilizer-state constructions over finite fields. The q-level states are built from known AME families over GF(q), and the qubit overlay exploits the product structure of the composite dimension 2q. The paper is firmly in the combinatorial-algebraic tradition of quantum information theory, not the variational or numerical-optimization camp. Results cascade outward from the nine-party case. The paper produces AME(11,2q) for q ≥ 5, AME(13,2q) for q ≥ 17, AME(15,2q) for q ≥ 9, and AME(17,2q) for q ≥ 67. For arbitrary odd N ≥ 3, the construction works whenever q ≥ max{N, 2^{N−2}}, meaning the required dimension of the q-level system grows exponentially in the party count N. Critically, Pandya proves this exponential growth is necessary within the coupling framework — it's not just a loose bound. The integrity regime here is mathematical proof, the gold standard. There's no simulation to interrogate, no benchmark to cherry-pick. The claims are either correct proofs or they aren't. The nine-party classification — AME(9,d) exists iff d ≥ 3 — combines these new constructions with previously established results and product operations. The paper acknowledges the exponential overhead honestly, marking it as a structural limitation rather than burying it. The milestone to watch is the party count frontier. Nine parties is now settled. Eleven through seventeen parties have constructions for large enough dimensions, but the full existence classification for 11 parties in every d remains open. The gap between 'exists for d = 2q with q large enough' and 'exists for every d ≥ k' is where the next work lives. The general bound q ≥ 2^{N−2} means the method gets expensive fast: for N = 21 parties, you'd need q ≥ 524288. A bonus result shows that for primes p ≡ 1 (mod 15), the AME(15,2p) states admit a cyclic symmetry — a shift of all 15 parties compensated by local unitaries. This is relevant for quantum error correction, where cyclic structure translates directly into simpler encoding circuits. The paper is 16 pages with one figure, pure math, no fluff.