What the LLM Should Not Say: Boundary-Aware Context Grounding for A Seven-Channel EEG Agent
Zhiyuan Xu, Yueqing Dai, Junling Li, Junwen Luo
Why It Matters
What makes this one worth your time
This work addresses the critical issue of misinterpretation in EEG analysis, which can lead to erroneous scientific conclusions, making it relevant for both researchers and practitioners in neuroscience and AI.
NeuraDock Agent enhances EEG data interpretation by grounding LLM responses in hardware and algorithmic context.
Summary
The paper presents the NeuraDock Agent, an architecture that integrates a deterministic EEG engine with a hardware-aware language layer to improve the reliability of interpretations from low-channel EEG data.
Key contributions
- Development of an open-source architecture that integrates a deterministic EEG engine with a language layer.
- Implementation of a boundary-aware context grounding mechanism that restricts LLM responses to validated information.
- Evaluation of the system through structured results and failure-injection experiments to ensure reliability.
Notable insights
- The separation of the EEG engine from the language layer allows for a more controlled interaction between the LLM and the EEG data, potentially reducing misinterpretations.
- The use of a boundary-awareness benchmark to evaluate the system's performance against adversarial questions is a novel approach to testing the robustness of LLMs in specialized domains.
Possible limitations
- Not stated in the abstract.
Abstract
arXiv:2606.26519v2 Announce Type: replace 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.