Built on Gemini 3

Neural Pattern Translator

A reasoning interface that interprets EEG signals into structured cognitive hypotheses — with evidence, alternatives, and explicit uncertainty.

Select an EEG recording to analyze

System Architecture

EEG RecordingMNE Sample Dataset60 channels · 4 conditionsauditory + visual stimuliSignal ProcessingMNE-PythonERP extractionBand power (δ θ α β γ)Electrode mappingGemini 3 — Pass 1Structured ReasoningHypotheses + evidenceUncertainty quantificationGemini 3 — Pass 2Narrative SynthesisPlain-language interpretationReasoning OutputDetected patternsRanked hypothesesRejected alternativesLimitationsNarrative explanationInterfaceStaged revealInteractiveplots + textInputClassical DSPAI Reasoning LayerOutputTwo-pass architecture: structured analysis → narrative synthesis

Live Neural Monitor

Simulated 10-channel EEG stream. Freeze at any moment to let Gemini 3 analyze the captured segment in real time.

EEG MONITOR10ch · 500Hz

Try It Yourself

Analyze your own EEG data

Upload a CSV file with channels as columns and time samples as rows. Values should be in microvolts (µV).

Drop a CSV file here or click to browse

.csv with channel columns × time rows

Expected CSV format
Fp1,Fp2,F3,F4,C3,C4,P3,P4,O1,O2
2.31,-1.05,0.87,-0.44,1.22,-0.98,0.15,0.67,-1.33,0.89
1.98,-0.77,1.12,-0.33,0.95,-1.11,0.28,0.54,-1.21,0.76
...
MNE-Python signal processingTwo-pass Gemini 3 reasoningReal EEG data