Bell's Theorem started with two qubits, spins 1/2. It is a no-go statement on classical (local causal) models of quantum correlations. Only after 25 years, it turned out that for three qubits the situation is even more mind boggling. General statements concerning higher dimensional systems, qutrits etc., started to appear even later, once the spin (higher than 1/2) picture was replaced by a more broad one, allowing all possible observables. This work is continuation of the Gdansk effort to take

more »

... advantage of the fact that Bell Theorem can be put in form of a linear programming problem, which in turn can be translated into a computer code. Our results are numerical and classify the strength of non-classicality of various families of three-qutrit states, as measured by resistance to noise. This is a previously uncharted territory. The results may be helpful in suggesting which three-qutrit states are handy for applications in quantum information protocols. One of the surprises is that the W state turns out to have a stronger non-classicality than the GHZ state.