The Mermin-Peres Magic Square
A three-by-three grid of two-qubit observables gives a compact, state-independent proof of quantum contextuality.
A grid that cannot be filled classically
The Mermin-Peres magic square is a 3x3 array of two-qubit observables, each built from Pauli operators, arranged so that the observables in every row and every column mutually commute and can be measured together. Quantum mechanics fixes the product of each row and column. Five of the six products equal +I and one equals -I, a pattern that no assignment of fixed plus or minus one values to the nine cells can reproduce.
The contradiction
Suppose each cell had a predetermined value of plus or minus one. The product of all three row-products would equal the product of all nine values, and likewise the product of all three column-products would equal the same product of all nine values. So row-products and column-products must multiply to the same thing. But quantum mechanics makes the three row-products multiply to +1 and the three column-products multiply to -1. This is an outright contradiction, independent of the state.
The array of individual signs cannot be consistently completed: the mismatched overall products (rows to +1, columns to -1) leave no valid classical assignment.
Why it is powerful
The magic square is state-independent: the contradiction holds for every quantum state, so no clever preparation is needed. It uses only two qubits and nine standard Pauli observables, making it experimentally accessible. It is a sharper, more economical demonstration of Kochen-Specker contextuality than the original ray colorings.
Uses
Beyond foundations, the magic square defines a nonlocal game, the Mermin-Peres game, that two entangled players win with certainty but classical players cannot. This connects contextuality to quantum advantage in games and communication, to self-testing of entangled states and measurements, and to the broader program of identifying contextuality as the resource behind quantum computational power.