CogEEGAgent: Toward Autonomous Cognitive EEG Analysis with Grounded Execution and Selection-Aware Verification

2026-07-27

AI Summary

This work addresses the challenge that cognitive EEG analysis heavily relies on expert knowledge, while existing large language model (LLM) agents often fail to align with user intent and lack verifiability. To overcome this, the authors propose a “semantic-scientific” dual-authority architecture: the LLM interprets user intent and generates a preregistered analysis plan, whereas a deterministic component enforces type-contract validation, access control, and evidence-bound result publication. This framework uniquely combines language-driven flexibility with fail-safe reasoning control, achieving the first auditable and bounded-autonomy automated analysis in cognitive EEG. Experiments demonstrate that the system outperforms deterministic routers on predefined routing benchmarks, accurately publishes supportive findings, blocks capability misuse and unauthorized reuse, and effectively suppresses false positives induced by adaptive search through its verification mechanism.

Abstract

Electroencephalography (EEG) analysis in cognitive studies requires specialized expertise and involves many defensible choices over contrasts, channels, time windows, and statistical tests. LLM agents can translate varied natural-language questions into analysis choices, offering a flexible interface for automation. Yet fluent reports alone cannot establish that an agent selected the requested analysis or evaluated a confirmatory claim independently of adaptive search. We present CogEEGAgent, a cognitive-EEG analysis agent grounded in MNE-Python. Its EEG-specific scientific harness separates semantic from scientific authority. The LLM interprets intent and proposes registered analyses, while deterministic components validate typed contracts, control confirmation access, and authorize evidence-bound release. On a prespecified routing benchmark, CogEEGAgent maps language to registered analyses more accurately than a matched deterministic router, while matched preflight makes both systems abstain whenever required. In an externally model-authored, outcome-blind campaign, the complete system releases supported analyses with participant-disjoint confirmation and blocks prespecified capability hazards and lifecycle-reuse requests. Policy stress testing shows that held-out confirmation curbs false positives from uncorrected adaptive search. Together, these studies establish bounded autonomy and an auditable automation framework for cognitive-EEG workflows. More broadly, they show how scientific agents can combine flexible language understanding with fail-closed control over inference and release.

Problem

• EEG analysis
• cognitive neuroscience
• autonomous agents
• scientific verification
• adaptive search

Innovation

• grounded execution
• selection-aware verification
• bounded autonomy
• scientific agent
• evidence-bound release