Exercises
LM-C12-C-001
If a dependency spans 10 tokens, is it longer than a bigram context? Answer 1
for yes or 0 for no.
Compute it first, then check your number.
LM-C12-C-002
If a signal is multiplied by 0.5 three times, what is the result times 8?
Compute it first, then check your number.
LM-C12-C-003
If a signal is multiplied by 3 twice, what is the result?
Compute it first, then check your number.
LM-C12-C-004
If hidden size is fixed at 256, how many numbers are in the hidden state after 100 time steps?
Compute it first, then check your number.
LM-C12-C-005
Was attention partly attractive because it gave a more direct route to earlier
positions? Answer 1 for yes or 0 for no.
Compute it first, then check your number.