PT-2026-59899

Impact If tf.raw ops.TensorListResize is given a nonscalar value for input size, it results CHECK fail which can be used to trigger a denial of service attack. python import numpy as np import tensorflow as tf a = data structures.tf tensor list newelements = tf.constantvalue=3, 4, 5 b = np.zeros0...

PT-2026-59853

Impact When RangeSize receives values that do not fit into an int64 t, it crashes. cpp auto size = std::is integral::value ? Eigen::numext::abslimit - start + Eigen::numext::absdelta - T1 / Eigen::numext::absdelta : Eigen::numext::ceil Eigen::numext::abslimit - start / delta; // This check does n...

PT-2026-59782

Impact If FractionMaxPoolGrad is given outsize inputs row pooling sequence and col pooling sequence, TensorFlow will crash. python import tensorflow as tf tf.raw ops.FractionMaxPoolGrad orig input = 1, 1, 1, 1, 1, orig output = 1, 1, 1, out backprop = 3, 3, 6, row pooling sequence = -0x4000000, 1...

PT-2026-59844

Impact The implementation of Conv2DBackpropInput requires input sizes to be 4-dimensional. Otherwise, it gives a CHECK failure which can be used to trigger a denial of service attack: python import tensorflow as tf strides = 1, 1, 1, 1 padding = "SAME" use cudnn on gpu = True explicit paddings =...

PT-2026-59980

Impact If the stride and window size are not positive for tf.raw ops.AvgPoolGrad, it can give an FPE. python import tensorflow as tf import numpy as np @tf.functionjit compile=True def test: y = tf.raw ops.AvgPoolGradorig input shape=1,0,0,0, grad=0.39117979, ksize=1,0,0,0, strides=1,0,0,0,...

PT-2026-59760

Impact The function tf.raw ops.LookupTableImportV2 cannot handle scalars in the values parameter and gives an NPE. python import tensorflow as tf v = tf.Variable1 @tf.functionjit compile=True def test: func = tf.raw ops.LookupTableImportV2 para='table handle': v.handle,'keys': 62.98910140991211,...

PT-2026-59908

Impact When running with XLA, tf.raw ops.Bincount segfaults when given a parameter weights that is neither the same shape as parameter arr nor a length-0 tensor. python import tensorflow as tf func = tf.raw ops.Bincount para='arr': 6, 'size': 804, 'weights': 52, 351 @tf.functionjit compile=True d...

PT-2026-59896

Impact FPE in TensorListSplit with XLA python import tensorflow as tf func = tf.raw ops.TensorListSplit para = 'tensor': 1, 'element shape': -1, 'lengths': 0 @tf.functionjit compile=True def fuzz jit: y = funcpara return y printfuzz jit Patches We have patched the issue in GitHub commit...

PT-2026-59759

Impact If the parameter indices for DynamicStitch does not match the shape of the parameter data, it can trigger an stack OOB read. python import tensorflow as tf func = tf.raw ops.DynamicStitch para='indices': 0xdeadbeef, 405, 519, 758, 1015, 'data': 110.27793884277344, 120.29475402832031,...

PT-2026-59981

Impact When RangeSize receives values that do not fit into an int64 t, it crashes. cpp auto size = std::is integral::value ? Eigen::numext::abslimit - start + Eigen::numext::absdelta - T1 / Eigen::numext::absdelta : Eigen::numext::ceil Eigen::numext::abslimit - start / delta; // This check does n...

PT-2026-59993

Impact If Requantize is given input min, input max, requested output min, requested output max tensors of a nonzero rank, it results in a segfault that can be used to trigger a denial of service attack. python import tensorflow as tf out type = tf.quint8 input = tf.constant1, shape=3,...

PT-2026-59995

Impact The RaggedRangOp function takes an argument limits that is eventually used to construct a TensorShape as an int64. If limits is a very large float, it can overflow when converted to an int64. This triggers an InvalidArgument but also throws an abort signal that crashes the program. python...

PT-2026-59778

Impact The function MakeGrapplerFunctionItem takes arguments that determine the sizes of inputs and outputs. If the inputs given are greater than or equal to the sizes of the outputs, an out-of-bounds memory read or a crash is triggered. Patches We have patched the issue in GitHub commit...

PT-2026-59811

Impact If BCast::ToShape is given input larger than an int32, it will crash, despite being supposed to handle up to an int64. An example can be seen in tf.experimental.numpy.outer by passing in large input to the input b. python import tensorflow as tf value = tf.constantshape=2, 1024, 1024, 1024...

PT-2026-59764

Impact If LRNGrad is given an output image input tensor that is not 4-D, it results in a CHECK fail that can be used to trigger a denial of service attack. python import tensorflow as tf depth radius = 1 bias = 1.59018219 alpha = 0.117728651 beta = 0.404427052 input grads =...

PT-2026-59773

Impact When tf.linalg.matrix rank receives an empty input a, the GPU kernel gives a CHECK fail that can be used to trigger a denial of service attack. python import tensorflow as tf a = tf.constant, shape=0, 1, 1, dtype=tf.float32 tf.linalg.matrix ranka=a Patches We have patched the issue in GitH...

PT-2026-59789

Impact When CollectiveGather receives an scalar input input, it gives a CHECK fails that can be used to trigger a denial of service attack. python import tensorflow as tf arg 0=1 arg 1=1 arg 2=1 arg 3=1 arg 4=3, 3,3 arg 5='auto' arg 6=0 arg 7='' tf.raw ops.CollectiveGatherinput=arg 0, group...

PT-2026-59786

Impact version:2.11.0 //core/ops/audio ops.cc:70 Status SpectrogramShapeFnInferenceContext c ShapeHandle input; TF RETURN IF ERRORc-WithRankc-input0, 2, &input; int32 t window size; TF RETURN IF ERRORc-GetAttr"window size", &window size; int32 t stride; TF RETURN IF ERRORc-GetAttr"stride", &strid...

PT-2026-59806

Impact NPE in RandomShuffle with XLA enable python import tensorflow as tf func = tf.raw ops.RandomShuffle para = 'value': 1e+20, 'seed': -4294967297, 'seed2': -2147483649 @tf.functionjit compile=True def test: y = funcpara return y test Patches We have patched the issue in GitHub commit...

PT-2026-59809

Impact If MirrorPadGrad is given outsize input paddings, TensorFlow will give a heap OOB error. python import tensorflow as tf tf.raw ops.MirrorPadGradinput=1, paddings=0x77f00000,0xa000000, mode = 'REFLECT' Patches We have patched the issue in GitHub commit...