Foundation models,
wired into your lab.
Cell, signal and connectome foundation models — same parameters, same result viewer, same audit trail as every other analysis on Neuron Lab.
Cell Atlas FM
PCA embedding · scGPT / Geneformer-class in production
Embed cells, then annotate cell types zero-shot or cluster them — accuracy on held-out cells.
Neural Signal FM
spectral + PCA · LaBraM / CBraMod-class in production
Embed neural-signal windows and classify dynamical state — regular / bursting / irregular.
Connectome FM
spectral node embedding · graph pretraining in production
Embed connectome nodes and detect communities; ARI vs planted modules.
Shared run pipeline
FM analyses go through the same launcher, storage, RLS and result viewer as every other module.
Embedding + metrics
Every FM returns a 2D embedding, quantitative metrics and per-class accuracy tables.
Drop-in backbones
PCA / spectral heads today; single load-weights hook to swap in scGPT, LaBraM, CBraMod.