.. meta:: :description: Backward pass of separable 3D max pooling that uses the input and output of 3D max pooling forward and the backward grad :date-modified: 10/19/2026 .. currentmodule:: nkilib.experimental.pooling Max Pooling 3D Backward Kernel API Reference ============================================ Backward pass of separable 3D max pooling that uses the input and output of 3D max pooling forward and the backward gradient. Background ----------- The ``max_pooling_3d_backward`` kernel computes the backward pass of 3D max pooling from the forward input, the forward output, and the output gradient. API Reference -------------- **Source code for this kernel API can be found at**: `max_pooling_3d_backward.py `_ max_pooling_3d_backward ^^^^^^^^^^^^^^^^^^^^^^^ .. py:function:: max_pooling_3d_backward(grad_output: nl.NkiTensor, input_tensor: nl.NkiTensor, output_tensor: nl.NkiTensor, kernel_size: Union[int, tuple], stride: Optional[Union[int, tuple]] = None, padding: Union[int, tuple] = 0, data_format: str = 'NCDHW', output_format: Optional[str] = None) -> nl.NkiTensor Backward pass of separable 3D max pooling that uses the input and output of 3D max pooling forward and the backward gradient. :param grad_output: Gradient w.r.t. the forward output, in HBM, output_format layout, forward-output shape. :type grad_output: ``nl.NkiTensor`` :param input_tensor: The forward input, data_format layout. :type input_tensor: ``nl.NkiTensor`` :param output_tensor: The forward output (per-window maxima), output_format layout, same shape as grad_output. :type output_tensor: ``nl.NkiTensor`` :param kernel_size: Pooling window (kD, kH, kW) (int broadcasts). :type kernel_size: ``Union[int, tuple]`` :param stride: Stride (sD, sH, sW); defaults to kernel_size. :type stride: ``Optional[Union[int, tuple]]`` :param padding: Implicit -inf padding (pD, pH, pW). :type padding: ``Union[int, tuple]`` :param data_format: Layout of input_tensor and grad_input, "NCDHW" or "NDHWC". :type data_format: ``str`` :param output_format: Layout of grad_output / output_tensor; defaults to data_format. :type output_format: ``Optional[str]`` :return: Gradient w.r.t. the forward input, in HBM, data_format layout, same spatial extents as input_tensor. :rtype: ``nl.NkiTensor`` **Dimensions**: * B: batch.