When Language Models Meet NeuroGraphs: Exploring Enhanced Agentic LLM Framework Towards Brain Network Analysis
2.60T1 sourcearXiv cs.MA
Source record
Published by arXiv cs.MA (T1 source). The original is at https://arxiv.org/abs/2607.22082.
Pipeline notes
The summary and note below are generated by the signal pipeline — they are Beyond Desk’s reading, not quotations from the source.
SummaryBrainAgent is an agentic LLM framework for brain network analysis that converts connectome data into multi-level structural descriptions, retrieves neuroscience knowledge and task cases, and iteratively reasons and reflects to produce verifiable classifications. It outperforms direct prompting and standard reasoning baselines on four rs-fMRI datasets.
Why it mattersA worked example of a retrieve-reason-reflect agent applied to scientific graph data, with ablations isolating each component's contribution. Useful reference for designing similar pipelines elsewhere.
Cited by
No citations on record.
