Human Mind

Measuring cognitive coherence from EEG brain activity. The same 13 cognitive functions measured in AI models are measured in human brains, enabling the first direct comparison of cognitive coherence across substrates.

What is measured

EEG recordings capture the electrical activity of the brain. The LOC framework maps frequency-domain power to the same 13 cognitive functions using a proprietary server-side algorithm. The SDK handles loading, preprocessing, and PSD extraction locally. Scoring is performed by the AIME API.

How to score an EEG recording

from aime_loc import LOC
from aime_loc.eeg import EEG

loc = LOC(api_key="sk-aime-...")
eeg = EEG(loc)

recording = eeg.load("subject01.set")
recording.preprocess()
epochs = recording.extract_epochs(duration=2.0)
profile = eeg.score(epochs, subject="sub-01")

print(profile.tc_score)        # 23.4
print(profile.best_function)   # Mindfulness
profile.radar_chart()

Supported formats

The SDK loads any EEG format supported by MNE-Python:

  • EEGLAB (.set)
  • EDF / EDF+ (.edf)
  • BrainVision (.vhdr)
  • BioSemi BDF (.bdf)
  • EGI MFF (.mff)
  • NumPy arrays and CSV files

Typical TC ranges

ConditionTypical TC Range
Resting eyes-closed25–40%
Resting eyes-open20–35%
Focused attention15–30%
Meditation30–50%
Sleep (N2)35–55%
Task switching10–20%

EEG TC scores are generally higher than LLM TC scores because brain activity naturally exhibits more hierarchical cognitive structure.

Data privacy

The SDK sends only Power Spectral Density (PSD) arrays to the API—frequency-domain aggregates that contain no temporal patterns, no raw waveforms, and cannot be reversed to identify individuals. A typical 30-minute recording produces ~650KB of PSD data.

Supported DevicesEEG Documentation