intermediate · Interactive lab
Entanglement
Two gloves in two boxes are correlated: open one and you know the other. An entangled pair agrees in every question you can ask it, and still carries no message. Telling those apart is the lesson.
By the end: Tell an entangled pair from a pre-agreed one by the only thing that separates them: agreement in more than one basis.
Your challenge
Start here. This lab opens with every claim marked as shown by the bench, which is the wrong starting point: two of these the bench actually refuses, one is settled by the theory rather than by any run, and one would need a real Bell test. Read the experiment on each row and judge it.
See it happen
Two pairs, two questions
- Pair source → The pairs
- The pairs → Both detectors: one question
- Both detectors → Tally
Two kinds of pair
The Pair source makes both: an entangled pair, and a pre-agreed pair carrying one bit chosen when it was made. Watch how often the two sides agree.
Learn more
Why this pattern exists
Put a left glove in one box and a right glove in another, send them to opposite ends of the country, and open one: you instantly know what is in the other. Nothing travelled, and nobody is surprised. That is a pre-agreed pair, and most explanations of entanglement are describing exactly this without noticing. The bench holds both kinds of pair at once. Ask both sides the Z question and they are identical: each pair agrees every single time. Ask both sides the X question instead, and the entangled pair still agrees every time while the pre-agreed pair drops to a coin toss. One bit decided in advance cannot be right in two different questions at once. And through all of it, each side on its own is a fair toss, whatever the far side is asked — which is why nothing here is a channel.
A source in the middle makes pairs and sends one side each way. Two kinds are on the bench: an entangled pair, and a pre-agreed pair carrying one bit decided when it was made. Both sides of each pair are measured in a basis you choose, and the tally counts how often the two sides agreed. Every graded experiment runs on a fixed seed.
- Tell an entangled pair from a pre-agreed one by the only thing that separates them: agreement in more than one basis.
- See why each side on its own carries no information, however strong the correlation between them.
- Say what this bench establishes, what the theory establishes, and what would need a real Bell test.
The rule this lesson applies: A pre-agreed pair carries one bit, written when the pair was made. It agrees perfectly in the basis that bit is written in, and is a coin toss in the other — because there is nothing there to agree about. An entangled pair agrees in both, and that is the observation a single shared bit cannot reproduce. Be careful about what follows. This bench refuses one particular classical model; it is not a Bell test, which uses measurement angles this apparatus does not offer, and which is where the impossibility of local hidden variables is actually established. What does follow, from the theory rather than from any run, is no-signalling: each side's own outcomes are 50/50 whatever the far side is asked, so the correlation cannot carry a message, and entanglement is not a faster-than-light channel. Its engineering uses are elsewhere — key distribution, where the correlation is used with a classical channel alongside it, and error correction, where entanglement across physical qubits is what makes a logical one. This page computes the statistics the standard model predicts; it models no noise, no decoherence and no hardware.