Loading demo links...
Showing 1–12 of 12 questions
What does a multimodal model process that a unimodal model does not?
Select an option, then click Submit answer.
Residual connections in deep networks primarily help with what problem?
Select an option, then click Submit answer.
A combined multimodal loss function typically does what?
Select an option, then click Submit answer.
Transfer learning between modalities is useful because it what?
Select an option, then click Submit answer.
Which architecture family underpins most vision-language models?
Select an option, then click Submit answer.
Training instability in multimodal models often arises from what?
Select an option, then click Submit answer.
Which TWO help stabilise multimodal training? (Choose two.)
Select all that apply, then click Submit answer.
Which TWO are deep learning fundamentals? (Choose two.)
Select all that apply, then click Submit answer.
DRAG DROP
Match each concept to its role.
Items to match: Encoder, Decoder, Residual connection, Loss function.
Answer is in the explanation below.
CORRECT TEXT
Connections that skip layers to ease gradient flow are called ______ connections.
Answer is in the explanation below.
Which TWO are analysis activities for multimodal data? (Choose two.)
Select all that apply, then click Submit answer.
DRAG DROP
Match each visualisation to its purpose.
Items to match: Attention map, Embedding projection, Confusion matrix, Output gallery.
Answer is in the explanation below.
