Measurement-Gated Provenance Attenuation for Frozen EEG Representations

2026-09-26

AI Summary

This study addresses the deep entanglement of device origin and neural activity in frozen EEG representations, where blind debiasing risks inadvertently removing genuine biological differences. We propose MGPA, establishing a “measurement defines scope, preservation defines target” principle. Specifically, it leverages measurement evidence to construct gating constraints and employs closed-form solutions alongside critic-guided iterative optimization to shift source scores toward preserved coordinate predictions at minimal cost, precisely disentangling measurement effects from biological signals without retraining. Controlled and cross-device experiments demonstrate that our method significantly reduces source predictability while maintaining or enhancing downstream task performance. By outperforming the LEACE baseline, this work validates the effectiveness of reusable representation correction for robust cross-device EEG analysis.

Abstract

Frozen EEG representations retain acquisition signatures as well as neural activity. Source predictability alone does not identify what should be removed: it can reflect measurement effects or genuine biological and population differences, which should not be erased. We propose Measurement-Gated Provenance Attenuation (MGPA), built on one principle: measurement evidence determines where correction may act, and preserved information determines what it should aim for. Paired measurement contrasts define a gate outside which nothing changes; inside it, the source score is moved to the value the preserved coordinates already predict: for a fixed affine score, this keeps the same information as any target set by those coordinates and needs the least expected squared movement. Closed-form and critic-guided iterative constructions apply it without source identity or encoder retraining. Three studies test the principle at increasing distance from its assumptions. Under controlled reference changes, where the source-task association is known, MGPA brings source to near chance with task performance unchanged, whereas erasing what predicts source (LEACE) lowers frozen-task AUROC from .753 to .656 while barely touching source; ablations attribute the attenuation to the gate's directions and 2.7x less movement to the conditional target. Across recordings from different devices and electrodes, iterative correction lowers source accessibility while preserving or improving task performance. Finally, one iterative map selected on one task and reused unchanged on existing heads for two others raises their worst-association AUROC (lowest over device-label shifts) by .057 and .019 over LEACE, at a cost to those heads while the training association holds. A reusable correction shows its value in how an existing predictor behaves once acquisition cues stop being reliable, not only in what a probe can read.

Problem

• EEG representations
• domain adaptation
• measurement effects
• provenance attenuation
• source predictability

Innovation

• Measurement-Gated Provenance Attenuation
• Frozen EEG Representations
• Domain Adaptation
• Artifact Removal
• Conditional Targeting