Exam

Guesses on what the content will be

Diagrams of architectures could be useful!

3 questions from Lech

• Interpretability

• Learning theory

• Stochastic neural networks?

• History?

• Learning?

3 questions from Ali

• Word2Vec vs onehot encoding? (Word representation)

• Recurrent models of language

• Sequence to sequence networks / attention

• Transformer