A unit’s receptive field is the region of the input image that can influence its activation. In AlexNet, deeper convolutional layers build on earlier features to capture patterns across broader image regions. section 2.5.1
AlexNet’s changing filters
AlexNet moves from 11×11 filters in early layers to 5×5 and then 3×3 filters deeper in the network. Alongside an increase in feature maps, this progression supports a shift from coarse features toward finer patterns while deeper units capture broader image regions. section 2.5.1
Figure 2.3 The AlexNet architecture (cropped in the original paper). The numbers indicate spatial dimensions and channel depth at each stage. The input image (224 × 224 × 3 RGB) passes through a series of convolutional and pooling layers that progressively reduce spatial resolution (224→55→27→13) while increasing channel depth. Early downsampling is driven by the first convolutional layer (11 × 11 filters with a stride of 4), followed by max pooling layers (3 × 3 windows with a stride of 2) that further reduce spatial dimensions. The stride determines how far a filter moves across the input at each step, controlling how densely the image is sampled and how quickly spatial resolution is reduced. Filter sizes (11, 5, 3) are depicted as cubes and define the local receptive field of each convolution. The two horizontal streams reflect the model’s original two-GPU training setup, where feature maps were split across devices (e.g., 48 + 48 = 96 channels). The fully connected layers (4096 neurons, implemented as 2048 + 2048 across GPUs) feed into a 1000-way output for ImageNet classification. Used with permission granted by the lead author (Alex Krizhevsky).
The architecture diagram shows filter sizes at successive stages, alongside the changes in spatial resolution and channel depth. section 2.5.1
