Click the round data qubits to inject errors. The diamond stabilizers light up whenever they detect that something nearby has flipped, without ever measuring the data itself. That detection pattern is the syndrome, and it's the only clue a real quantum computer gets. This is the foundation of how we protect quantum information.
A single quantum bit is fragile, the slightest noise flips it and your computation is ruined. The fix isn't a better qubit; it's spreading one piece of information across many physical qubits so that no single failure destroys it. That bundle is a logical qubit, and the layout above, the rotated surface code, is today's leading way to build one.
The trick is the diamond-shaped stabilizers. Each one constantly asks a yes/no question about its neighboring data qubits, but a question carefully designed so the answer reveals that an error occurred without revealing what the data is. (Measuring the data directly would destroy the superposition; that's the heart of the problem QEC solves.)
When a stabilizer's answer changes, it lights up. The full pattern of lit stabilizers is the syndrome, and there's a catch you can see above: an error chain only lights up the stabilizers at its two endpoints. From that sparse clue, a decoder has to guess what actually went wrong and undo it. That guessing is where most of the hard research lives, and it's what the next module lets you watch in action.
On accuracy: the lattice connectivity, stabilizer checks, and syndrome computation here are physically faithful to the rotated surface code. This page shows the detection step. Decoding, noise models, and threshold behavior are coming as separate, clearly-labeled modules, so nothing here is a cartoon standing in for real physics.