Issues

• Add an x64_mode setting for 64-bit dtype support (off / explicit_x64 / enable_x64)

  backend:jaxbackend:tensorflowbackend:torchgTech-team-review-pendingtype:feature
  Feature

  keras-team/keras#24040 · Filter by author gaga1313 opened on Oct 21, 2026
  

• Model config can execute code from the working directory via the domain-package import in _retrieve_class_or_fn

  gTech-team-review-pending
  keras-team/keras#24034 · Filter by author LinZiyuu opened on Oct 21, 2026
  1

  

• [Feature / Bug]: Missing @keras_export for functional cosine_similarity metric and docstring discrepancy

  gTech-team-review-pending
  keras-team/keras#24019 · Filter by author GauravPatil2515 opened on Oct 20, 2026
  

• keras.ops.image.resize no longer rejects a non-positive size given as NumPy integers

  gTech-team-review-pending
  keras-team/keras#24007 · Filter by author lequangdung2005 opened on Oct 20, 2026
  1

  

• ops.one_hot raises for indices >= num_classes on the torch backend but returns zeros on TensorFlow

  backend:tensorflowtype:Bug
  keras-team/keras#24006 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  1

  

• ops.arange raises on an empty range on TensorFlow and torch

  keras-team/keras#24005 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  2

  

• ops.split fails with negative or out-of-range indices_or_sections on TensorFlow and torch

  keras-team/keras#24004 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  2

  

• ops.diagonal, ops.diag, ops.trace and ops.eye raise for out-of-range offsets on the TensorFlow backend

  backend:tensorflowtype:Bug
  keras-team/keras#24003 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  1

  

• ops.tril / ops.triu raise for k at or beyond the matrix size on the TensorFlow backend

  keras-team/keras#24002 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  2

  

• ops.vstack of 1-D inputs returns a flat array on the TensorFlow backend

  backend:tensorflowgTech-team-review-pending
  keras-team/keras#24001 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  2

  

• ops.unfold, ops.unstack, ops.stop_gradient and ops.vectorized_map fail on non-tensor inputs with the torch backend

  type:Bug
  keras-team/keras#24000 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  1

  

• ops.image.gaussian_blur returns the wrong output shape for even or non-square kernels on the torch backend

  keras-team/keras#23999 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  2

  

• ops.image.gaussian_blur returns wrong values for non-square kernel_size on the TensorFlow backend

  keras-team/keras#23998 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  2

  

• ops.image.map_coordinates raises GatherNd out-of-range error with fill_mode="mirror"/"reflect" on the TensorFlow backend

  backend:tensorflowtype:Bug
  keras-team/keras#23997 · Filter by author MaddipatlaChetan24 opened on Oct 20, 2026
  1

  

• Remove TF dependencies in non-TF backends

  keras-team/keras#23996 · Filter by author SamanehSaadat opened on Oct 19, 2026
  

• Automate PR/Issue status label updates when contributors respond or push changes

  type:feature
  keras-team/keras#23990 · Filter by author pctablet505 opened on Oct 17, 2026
  

• Torch backend: ops.std and ops.logsumexp reduce over every axis when axis=()

  backend:torchgTech-team-review-pending
  keras-team/keras#23989 · Filter by author Anu27n opened on Oct 17, 2026
  2

  

• ops.heaviside returns 0 (Torch) or x2 (TensorFlow) for NaN input instead of NaN

  backend:tensorflowbackend:torchtype:Bug
  keras-team/keras#23988 · Filter by author Anu27n opened on Oct 17, 2026
  1

  

• TensorFlow backend: sort, argsort, argmax, argmin, unique and top_k mishandle NaN

  backend:tensorflowgTech-team-review-pending
  keras-team/keras#23987 · Filter by author Anu27n opened on Oct 17, 2026
  

• Move backend-independent symbols from backend to backend.common.

  keras-team/keras#23979 · Filter by author hertschuh opened on Oct 16, 2026
  

• Cross Backend Gradient Discrepancy in a Conv2D Softplus Dense Model

  backend:tensorflowbackend:torchlayersstat:awaiting response from contributor
  keras-team/keras#23973 · Filter by author L4AIsS opened on Oct 16, 2026
  1

  

• standardize_tuple returns unconverted floats/numpy ints despite validating them as int

  type:Bug
  keras-team/keras#23966 · Filter by author MaddipatlaChetan24 opened on Oct 15, 2026
  2

  

• Progbar.update formats negative metric values differently depending on verbosity

  type:Bug
  keras-team/keras#23962 · Filter by author MaddipatlaChetan24 opened on Oct 15, 2026
  2

  

• smart_resize’s rank-validation guard crashes on tensors of unknown rank

  backend:tensorflowtype:Bug
  keras-team/keras#23961 · Filter by author MaddipatlaChetan24 opened on Oct 15, 2026
  1

  

• standardize_tuple returns unconverted floats/numpy ints despite validating them as int

  gTech-team-review-pendinglayers
  keras-team/keras#23960 · Filter by author MaddipatlaChetan24 opened on Oct 15, 2026
  4