51. PyTorch: MaxPool2D Forward

Implement max pooling — the standard downsampling operation in CNNs. A sliding window takes the maximum value in each window, reducing spatial dimensions while retaining the strongest activations.

For each pooling window of size pool_size × pool_size, output the maximum value across all spatial positions (per channel).

Signature: def maxpool2d(x, pool_size=2, stride=2)

• x: (H, W, C) — input feature map, channels-last
• pool_size: int — pooling window size (default 2)
• stride: int — step between windows (default 2)
• Returns: output of shape (H_out, W_out, C) where H_out = (H - pool_size) // stride + 1.

The test harness checks the pooled output values.