Skip to main content
Theorem
Path
Curriculum
Atlas
Paths
Labs
Diagnostic
Evidence
Blog
Search
Sign in
Pulse
/
Transformers
Transformers Pulse
Five questions across 7 topics. No timer.
1 / 5
Transformers
Question 1 of 5
foundation (1/10)
conceptual
What happens during the forward pass of a neural network?
Hide and think first
A.
The model deletes activations because they are only needed during inference
B.
The test set is mixed into the batch so the model can generalize better
C.
The optimizer permanently changes the weights before seeing any data
D.
The model computes activations and logits from inputs using its current weights
Show Hint
Submit Answer
I don't know