Skip to content

Fix NaN in jax.nn.standardize due to negative variance #30426 #33496

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
wants to merge 4 commits into
base: main
Choose a base branch
from
+22 −1
Open

Fix NaN in jax.nn.standardize due to negative variance #30426 #33496

Changes from 1 commit
Commits
Show all changes
4 commits
Select commit Hold shift + click to select a range
b7b33df
Fix NaN in jax.nn.standardize due to negative variance #30426
nandan2003 Nov 24, 2025
2d47d3b
Update jax/_src/nn/functions.py
nandan2003 Nov 24, 2025
4b79206
fix(nn): use lax.abs in standardize to preserve gradients
nandan2003 Nov 25, 2025
cc7fb77
test(nn): add regression and gradient stability tests for standardize
nandan2003 Nov 25, 2025
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 1 addition & 1 deletion jax/_src/nn/functions.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -693,7 +693,7 @@ def standardize(x: ArrayLike,
# when used in neural network normalization layers
variance = jnp.mean(
jnp.square(x), axis, keepdims=True, where=where) - jnp.square(mean)
return jnp.subtract(x, jnp.asarray(mean)) * lax.rsqrt(jnp.asarray(variance) + epsilon)
return jnp.subtract(x, jnp.asarray(mean)) * lax.rsqrt(jnp.maximum(jnp.asarray(variance), epsilon))

# TODO(slebedev): Change the type of `x` to `ArrayLike`.
@api.jit(static_argnames=("num_classes", "dtype", "axis"))
Expand Down