Neural network layer
A set of units that transforms an input vector into an output vector via a learned weight matrix, stacked with other layers to build a network. Each layer's output shape must match the next layer's expected input shape, which is the single most common source of a runtime error when building one in code.
Why exams ask this
Tested as a shape-mismatch debugging scenario in code: the exam gives a stack trace or a dimension error and wants the candidate to identify which layer's input or output shape does not match its neighbor.
Relevant to
Related concepts
Resources
No resources linked to this concept yet.