N Neurarch All checks Caught bugs Docs Open the app

Checks / pattern

Sigmoid / Tanh in deep networks

Check R11. Runs in the editor as you build, in CI through the GitHub Action, and over the wire at POST /api/v1/check. Milliseconds, before any GPU is billed.

info pattern R11
TriggerSigmoid or Tanh activation appears in a network with ≥ 5 weight-carrying layers.
WhySaturating activations vanish gradients in deep stacks. ReLU / GELU / SiLU are the modern defaults.
SourceGlorot & Bengio 2010 on the vanishing gradient problem.

Why it is not a lint you can ignore

A structural mistake does not fail at review time and it does not fail at import time. It fails when the module is constructed on the training node, after the job was queued and the dataset was downloaded. That is why this runs before the spend and not after it.
Run this check on your own model Free, no account needed

Every check

41 structural checks: 6 guardrail gates and 35 architecture advisor rules. See the full catalogue.

← R10 Attention without positional encoding  ยท  R12 Deep network with no normalization →