Issues · tensorflow/tensorflow · GitHub

Labels Milestones New issue

Question about documentation

#119862

· deathlegionteamlk opened

on May 31, 2026

tf.nn.relu6 under CPU XLA drops NaN and returns 0, while eager/autocluster propagate NaN

#119628

· MRiffiAslett opened

on May 27, 2026

`tf.image.adjust_hue` has inconsistent NaN propagation between eager and XLA

#119607

· rookieLiu2018 opened

on May 27, 2026

[DOCS] Two stale README contents

#119584

· doehyunbaek opened

on May 27, 2026

`tf.linalg.det` gradient crashes on singular matrices

#119477

· wuyii8941 opened

on May 26, 2026

`tf.math.xlogy` and `tf.math.xlog1py` gradient w.r.t. x is incorrectly zero when x=0

#119476

· wuyii8941 opened

on May 26, 2026

TFLite mnist_tflite tutorial is not runnable as written

awaiting PR merge

#119439

· fallintoplace opened

on May 25, 2026

XLA `tf.nn.top_k` returns different results from eager execution on NaN input

#119438

· MRiffiAslett opened

on May 25, 2026

`tf.math.cumprod` gradient is incorrect when input contains zeros

#119430

· wuyii8941 opened

on May 25, 2026

Grappler ArithmeticOptimizer incorrectly eliminates `reciprocal(reciprocal(x))` as involution

#119429

· wuyii8941 opened

on May 25, 2026

Grappler constant folding optimizes `x * 0` → `0`, suppressing NaN for Inf/NaN inputs

#119428

· wuyii8941 opened

on May 25, 2026

[TFLite] FULLY_CONNECTED reference kernel should use MultiplyByQuantizedMultiplier instead of std::round for requantization

#119412

· KavinSai-synthara opened

on May 25, 2026