Boundary-Aware Context Grounding for A Low-Channel EEG Agent

2026-06-24

AI Summary

This work addresses the challenge of spurious model interpretations arising from low-channel electroencephalography (EEG) data, which suffers from spatial sparsity and high signal variability. To mitigate this, the authors propose the NeuraDock Agent architecture, which decouples a deterministic local EEG processing engine from a hardware-aware language layer and enforces strict input constraints through a context-bundle mechanism. This mechanism restricts the large language model to only access vetted information regarding hardware specifications, algorithmic parameters, and scientific boundaries. By integrating local data parsing, quality control, and spectral analysis—and leveraging versioned context bundles with allow-list summaries—the system ensures reliable grounding of model outputs. Empirical evaluation demonstrates consistent results across ten repetitions on 12 EEG recordings, validates boundary enforcement in 288 question-answer pairs, and confirms robust retention of local artifacts even under anomalous conditions such as HTTP failures.

Abstract

Large language models (LLMs) can make scientific software easier to use. However, a general model does not automatically know which measurements a particular sensor can support, which algorithms are implemented in the current software, or which conclusions are justified by a computed result. These distinctions are especially important for low-channel electroencephalography (EEG), where sparse spatial coverage and variable signal quality make plausible but unsupported interpretations easy to produce. We present NeuraDock Agent, an open-source architecture that separates a deterministic local EEG engine from a hardware-aware language layer. The numerical engine parses recordings, performs quality control, executes reviewed spectral workflows, and writes machine-readable artifacts. The LLM receives only a compact, allowlisted summary and a versioned context pack. The context describes the seven-channel hardware, reviewed workflows, result fields, implementation boundaries, scientific limits, and reference cases. Raw EEG and dense per-sample arrays remain local We evaluate the system at three levels. First, 12 recordings produced identical structured results over ten numerical repetitions, and a complete Rest/Task run produced identical result, report, and figure hashes over three repetitions. Second, request-capture and failure-injection experiments confirmed the tested data boundary and preservation of local artifacts under HTTP, malformed-output, and connection failures. Third, a boundary-awareness benchmark tested 36 ordinary and adversarial questions under four context ablations and two LLMs, yielding 288 outputs.These results support hardware- and implementation-aware grounding as a practical mechanism for calibrating what an EEG agent accepts, qualifies, or refuses; they do not establish clinical validity or a validated absolute cognitive-load index.

Problem

• low-channel EEG
• boundary-aware grounding
• context grounding
• large language models
• scientific software

Innovation

• boundary-aware grounding
• low-channel EEG
• hardware-aware LLM
• deterministic EEG engine
• context-controlled agent