Research · NeuroAI & people
- machine-learning
The One Experiment Where the Split Didn't Matter
Every EEG and wearable experiment in this thread showed a leakage gap. Then I tried the same question on education data, and the gap nearly disappeared — which turned out to be the interesting part.
Read the article - eeg
Reproducing a Working-Memory Figure — and What Didn't Match
I tried to reproduce a published EEG finding on public data. My version came back the opposite direction. Here's the attempt, the honest result, and why I'm not calling it a refutation.
Read the article - eeg
What I'd Check Before Believing Any Brain-Decoding Result
Six small experiments in, here is the checklist I now run against any EEG or wearable-sensor accuracy claim — including my own.
Read the article - machine-learning
Wrist vs. Chest — How Much Does Sensor Placement Change the Answer?
Same people, same task, same evaluation discipline — just a comfortable wrist band instead of a clinical chest rig. What does that actually cost, in numbers?
Read the article - eeg
How Much Data From One New Person Does a Model Need?
Starting from a model that has never seen you, how many of your own trials does it take to start closing the gap? A small calibration-curve experiment on real motor-imagery EEG.
Read the article - eeg
The Same Model, Scored Twice — Random Split vs. Subject-Wise Split
I trained the same classifier on the same EEG data and scored it two different ways. One score was a lie. Here is the experiment, the numbers, and where I got the first version wrong.
Read the article - eeg
A Boring Baseline That Is Hard to Beat (CSP + LDA)
The standard motor-imagery pipeline, on one subject, with the honest number — weak, noisy, and still useful as a reference point for everything built on top of it.
Read the article - eeg
Cleaning a Brain Signal: What Filtering Changes, and What It Hides
A band-pass filter and a notch filter, applied to one real EEG recording, with the before-and-after spectrum to show exactly what they removed.
Read the article - eeg
What I Found When I Opened My First EEG File
Before any model, before any question about accuracy — what is actually inside a single real EEG recording, and what did I get wrong reading it the first time?
Read the article - eeg
What Does an EEG Recording Actually Look Like to a Computer?
A computer science perspective on EEG data—from raw voltages to processed signals—and why preprocessing decisions matter more than the model itself.
Read the article