A denoiser measured on people it never saw.

Before any two recordings can be compared, the artifacts have to come out the same way for everyone.

The programme's long-term interest is comparing whole-body recordings across people, devices and eventually clinical groups. Any such comparison needs a cleaning step that works on people it was not trained on, does not depend on the electrode layout, and does not erase the rhythms the comparison is about. That is a testable question with public data, so it was tested first, while the gait side of the project waits for a dataset it can legitimately use.

The study was also a chance to apply the workspace's evidence rules to a real result: person-level splits, recorded manifests and hashes, classical baselines, per-person reporting, a downstream check, and an immutable run bundle.

Seven steps, all of them reproducible from a command list.

Twice the filter's gain, and the gain is where filters are powerless.

On 20,000 clean windows corrupted with the training mixture (input signal-to-noise ratio about −5 dB), the band-pass filter recovered +7.1 dB and EEG-Clear +14.4 dB. The relative error fell from 1.27 (filter) to 0.44, and the correlation with the true signal rose from 0.68 to 0.86. The advantage is concentrated on blinks, eye movements, electrode pops and in-band muscle activity, the artifacts that share their frequencies with the brain signal; it disappears, as it should, for mild slow drift, where a high-pass filter is already the right tool.

Means of per-person means over five held-out people; the interval is a 2,000-sample subject bootstrap.

Clean input passed through with a correlation of 0.991 to the target, against 0.985 for the filter. On 1,500 real frontal windows containing genuine blinks of several hundred microvolts, for which no clean reference exists, the model removed 79 percent of the power that the filter left in place while keeping the rhythm on either side of each blink. Inference runs at 2,500 to 4,800 two-second windows per second on an eight-thread laptop CPU.

Waveform accuracy did not, by itself, preserve the movement rhythms.

A downstream probe took cue-locked epochs from the real left-fist and right-fist runs of each held-out person, computed mu (8 to 13 Hz) and beta (13 to 30 Hz) band power over seventeen sensorimotor electrodes, and asked two questions: how well do those features agree with the clean epochs, and can a classifier trained on clean epochs still tell left from right?

The first-stage model, trained on squared error alone, scored superbly on every waveform metric and yet made the band-power features of moderately contaminated epochs less faithful than leaving them alone (correlation 0.57 against 0.83 for the untouched epochs). Squared error rewards hedging: when the network cannot tell rhythm from artifact it shrinks both, which costs little signal-to-noise ratio but shifts every band-power feature by about the amount that separates one condition from another.

A second stage that added a band-power term to the loss raised that fidelity to 0.70, restored features that severe contamination had destroyed (0.70 against 0.52 for doing nothing), brought decoding accuracy from 0.53 back to 0.59 against a clean 0.60, and improved every other metric as well. The remaining gap at moderate contamination means that windows a screen already accepts as clean are best left alone: the rule that follows is to clean only what fails a screen, and to measure spectral fidelity explicitly whenever a cleaning step precedes a comparison across people.

A linear projection beat a neural network, and the label never needed either.

The workspace also held a two-person table of 2,548 pre-computed EEG statistics with an emotion label, no waveforms and no participant identifiers. The same corrupt-and-restore idea was applied to it in feature space. A principal-component projection recovered 63 to 69 percent of the lost feature variance at moderate and strong corruption; a residual autoencoder trained on the 1,494 available rows recovered almost none. The emotion label was indifferent to all of it, scoring 97 percent whether the features were clean, corrupted or restored, because 2,548 columns encode the same three classes redundantly. The table cannot say anything about cleaning EEG signals, because it contains none, and it remains ineligible for every Norysr tier.

Read the numbers with their boundary attached.

• The artifacts were synthetic; real ones were only inspected visually, and an artifact of an unmodelled shape may be handled worse.
• Every recording came from one laboratory amplifier and healthy adults. The tool accepts other devices, but the numbers do not cover them, children, or anyone with cerebral palsy.
• Five test people give wide intervals; one random seed was used; the clean-window rule was set after inspecting three training people.
• Classification accuracy in the probe was near chance even on clean epochs, so only its fidelity measure carried statistical weight.
• Seeing one channel at a time forgoes the spatial information that multi-channel methods use; it was chosen for device independence.
• It is not a Norysr device result, it did not use Norysr hardware or participants, and it changes nothing about the gait programme's stage gates.

Every number traces to a sealed run bundle.