True/False

Phoneme representations can help a learning algorithm understand basic sound components and thereby improve its performance.