# In the time dimension

# Weight the sum of the past

# To predict the future

mask = jnp.where(causal, ,0)

# Incorrect answer

# Try again