One of the prototype's success criteria is simple to state and easy to check: with eyes closed, occipital alpha (~10 Hz) at Oz should rise relative to an eyes-open block. The verifier does not interpret this — it just makes the contrast visible in the power spectrum so a human can judge it.
What the figure shows
The PSD plot overlays an eyes-open and an eyes-closed segment for the Oz channel. A clear bump near 10 Hz in the eyes-closed trace, sitting above the 1/f background, is the signal we expect. No bump, or a bump in the wrong channel, is a prompt to check electrode contact and reference before reading anything further.
Why we don't over-claim
A visible alpha increase says the montage and reference are behaving as intended for a healthy resting contrast. It does not say anything clinical, and it does not decode a mental state. The four-channel montage is a hard constraint on any spatial inference, and we keep it that way on purpose.