Claude mlp_and_activations — practice questions
15 free MCQs with worked solutions. Tap any question for the answer + explanation, or practice them all in the app.
Practice Claude mlp_and_activations in the app →What does an MLP add to a linear model?Why does a stack of affine layers collapse without an activation function?The ReLU activation function is defined asThe sigmoid function squashes its input to which interval?The tanh function squashes its input to which interval?The derivative of the ReLU function isThe derivative of sigmoid(x) equalsThe universal approximation theorem (Cybenko 1989) for MLPs states thatWhy has ReLU largely replaced sigmoid in hidden layers?Parametrised ReLU (pReLU) adds which capability over plain ReLU?For an MLP layer with input X ∈ ℝⁿˣᵈ, hidden width h and output q, the weight shapes areWhy is the sigmoid prone to the vanishing-gradient problem?Depth often beats width for the same expressive power becauseA 'thresholding' activation taking 0 or 1 cleanly is impractical for gradient-based learning becauseWhy are sigmoids still used in recurrent networks (e.g., LSTMs) even after ReLU's rise?