StackedML
Practice
Labs
Questions
Models
Pricing
Sign in
Questions
/
Deep Learning
/
Regularization & Stability
/
Dropout
← Previous
Next →
242.
Dropout Mechanism
easy
What does dropout do during the training forward pass?
A
It randomly adds Gaussian noise to each neuron's activation, smoothing the gradient signal during backpropagation
B
It randomly multiplies each neuron's activation by a value sampled from a uniform distribution between 0 and 1
C
It randomly swaps activations between pairs of neurons, forcing the network to learn permutation-invariant representations
D
It randomly sets each neuron's activation to zero with probability p, effectively removing it from the network for that step
Sign in to verify your answer
← Back to Questions