The Notebook — Vol. 01Methods · Tooling · EssaysUpdated 2026

Writing & technical notes

Notes from the seam between the lab and the model.

Short write-ups on methods, tooling, and the gotchas that come from working with real, noisy, motion-corrupted data — plus longer essays on neuroscience and machine learning. [ placeholder entries — replace with real posts ]

I.Technical Notes03 entries
N01Decoding millisecond EEG through motion artifacts A practical pipeline for recovering fast neural dynamics from movement-corrupted recordings — windowing, artifact rejection, and what actually survives.EEGSignal Processing20258 min readN02From 3 weeks to 2 days: parallelizing preprocessing Restructuring a multivariate neuroimaging pipeline for HPC clusters without losing reproducibility.HPCPython202411 min readN03Graph features from functional connectomes Extracting network statistics that actually predict clinical outcomes in TBI recovery.fMRIMachine Learning20249 min read
II.Essayson the blog
E01What neuroscience taught me about machine learning [ Essay placeholder — pull the real post from the bearblog. ]NeuroscienceAIBlogbearblog ↗

More longer-form writing lives on my bearblog.

Essays and notes on neuroscience, AI, and machine learning.

Visit the blog