C
CTX Neural·
U
User
Workspace
  • Workspace
  • Score creative
  • Generate
  • Performance
  • Competitive intel
  • Connections
Reference
  • Recommendations
  • Methodology
  • Settings
Cell coverage
N=4 / 50

Predictions return as Directional until N≥50 in this cell.

WorkspaceScoreAnalyticsConnectTips
How predictions work

Methodology

Honest about what we predict, what we don't, and why our confidence intervals exist.

vmPFC78
NAcc65
DMN72
Amygdala58
Mentalizing70

The model stack

Every creative passes through three layers. Layer 1 (frozen): a vision + audio + language encoder predicts a per-region brain time-series. We never modify these weights. Layer 2 (learned): a calibration head per (channel × format × vertical) cell maps brain features + multimodal sidecar features to your business outcome. Layer 3 (statistical): a split-conformal wrapper produces a per-prediction confidence interval calibrated to actual outcomes in your cell.

The model gets sharper as your calibration set grows. Wider CIs at low N are honest uncertainty, not a defect.

Signals: replicated vs dropped

Used in predictions (5 signals)
Gamma power
Cortical
Only signal to replicate across all 5 datasets (n=188)
Strong
vmPFC AUC
vmPFC
Valuation signal. Genevsky 2025 generalizes to non-representative samples
Strong
NAcc peak
NAcc
Anticipated reward. Aggregate-level only — not per-individual
Strong
DMN engagement window
DMN
Narrative self-relevance. Strong for story-driven formats
Strong
Mentalizing-network coupling
TPJ / MPFC
Social cognition, trust. Health domain MPFC findings — applied by analogy to D2C
Moderate
Excluded — failed replication (5 signals)
Alpha asymmetry
Frontal EEG
Failed replication across all 5 datasets. Widely used by EEG vendors — not valid.
Dropped
Inter-subject correlation (ISC)
Cortical
Did not replicate as a commercial predictor. Neural synchrony ≠ purchase intent.
Dropped
Theta power
Frontal EEG
Failed replication in n=188 k=161 systematic review.
Dropped
Beta power
Motor / EEG
Failed replication. No commercial predictive value found.
Dropped
Per-individual purchase prediction
vmPFC
Generalizes to population aggregate, not individual. Misuse is a major vendor tell.
Dropped

Most neuromarketing vendors quietly include the dropped signals. We exclude them because the Boksem 2025 systematic review (n=188, k=161 trailers, 5 datasets) is the most rigorous test to date. This is the wedge — believing the replication literature.

What we don't claim

This is not B2B-validated. It's applied to B2B based on adjacent-domain evidence.
Predictions are predicted, not measured. We never claim to read minds.
Outputs are aggregate-level. Per-individual purchase prediction does not generalize.
Mental health vertical has additional ethical guardrails — recommendations route through a safety filter.