I Entangled, mixed state with conditional entropy zero

Click For Summary
Classical conditional entropy is always non-negative, but in quantum mechanics, it can be negative, particularly in entangled states. The discussion explores the possibility of constructing an entangled, mixed bipartite state where conditional entropy is exactly zero. This scenario raises questions about the implications of zero conditional entropy in quantum systems, contrasting with classical probability where total system entropy is always greater than or equal to the entropy of its parts. In quantum mechanics, maximum knowledge of the entire system can coexist with less knowledge of its subsystems, leading to unique entropy relationships. The vanishing of conditional entropy indicates equal entropies without any additional special significance.
greypilgrim
Messages
581
Reaction score
44
Hi.
The classical (Shannon) conditional entropy is never negative. It can be written as ##H(Y|X)=H(X,Y)-H(X)## which allows for a quantum generalization using von Neumann entropy. In the case of entangled states, it can become negative.

I guess it should be possible to construct an entangled, mixed (bipartite) state where ##H(Y|X)## is exactly zero (though I don't know how to exactly do that). Does this have a specific meaning?
 
Physics news on Phys.org
Conditional entropy in classical probability is always positive because the entropy of the total system is always greater than or equal to the entropy of its parts. This is no longer true in QM and it can be negative because you can have maximum knowledge of the whole system (it's in a pure state), but less than optimal knowledge of the parts (subsystems are in a mixed state). It vanishing is just the special case where the entropies happen to be equal, it has no additional special meaning.
 
  • Like
Likes Greg Bernhardt and vanhees71
We often see discussions about what QM and QFT mean, but hardly anything on just how fundamental they are to much of physics. To rectify that, see the following; https://www.cambridge.org/engage/api-gateway/coe/assets/orp/resource/item/66a6a6005101a2ffa86cdd48/original/a-derivation-of-maxwell-s-equations-from-first-principles.pdf 'Somewhat magically, if one then applies local gauge invariance to the Dirac Lagrangian, a field appears, and from this field it is possible to derive Maxwell’s...