Skill UI

NeMo Curator streams GPU-accelerated curation for LLM training, with super-fast fuzzy deduplication, multimodal quality filters, semantic dedupl, PII redaction, NSFW checks, and RAPIDS-scaled pipelines for prepping massive corpora.

End-to-end Neuropixels recording analysis covering SpikeGLX/OpenEphys/NWB loading, preprocessing, motion correction, Kilosort4 sorting, quality metrics, Allen/IBL curation, and AI-assisted visualization for extracellular electrophysiology.

A comprehensive end-to-end toolkit for analyzing high-density Neuropixels extracellular recordings. It guides the user through the complete electrophysiology workflow, including data loading (SpikeGLX, NWB), robust preprocessing, motion/drift correction, advanced spike sorting (Kilosort4, etc.), quality metric calculation, and AI-assisted unit curation. Ideal for generating publication-ready results from large neural datasets.

This skill provides robust functionality to resolve free-text scientific labels into standard ontology term IDs (CURIEs) and validate existing identifiers against authoritative databases. It is essential for metadata curation, ensuring that submitted biological data (e.g., tissue types, diseases, assays) uses current, correct, and non-obsolete identifiers, thereby maintaining high data quality for large-scale scientific datasets.