Deep Learning Recurrent Neural Networks 2

Question 1

What is a bidirectional RNN?

Accepted Answer

An RNN that processes the sequence in both forward and backward directions simultaneously

Answer

A bidirectional RNN runs two RNNs on the input — one forward and one backward — and concatenates their outputs, giving each step access to both past and future context.

Question 2

In sequence-to-sequence (Seq2Seq) models, what is the role of the encoder?

Accepted Answer

Compressing the input sequence into a fixed-length context vector

Answer

The encoder processes the input sequence and compresses it into a context vector (the final hidden state) that summarizes the input for the decoder.

Question 3

What is teacher forcing in RNN training?

Accepted Answer

Feeding the ground-truth previous output as the next input during training rather than the model's own prediction

Answer

Teacher forcing uses the ground-truth token at each step as the next input during training, stabilizing learning by avoiding compounding errors from the model's own predictions.

Question 4

What is gradient clipping used for when training RNNs?

Accepted Answer

Capping gradient magnitudes to prevent exploding gradients from destabilizing training

Answer

Gradient clipping rescales gradients when their norm exceeds a threshold, preventing the large weight updates that cause divergence due to exploding gradients in deep or long RNNs.

Question 5

What type of tasks are RNNs / LSTMs most naturally suited for?

Accepted Answer

Sequential and time-series tasks such as language modeling, speech recognition, and forecasting

Answer

RNNs and LSTMs are designed for sequential data where order matters, making them well-suited for text, audio, time series, and other temporal tasks.

Deep Learning Practice Test

Deep Learning Practice Test

Deep Learning Recurrent Neural Networks 2