DLNLP-5 QA&Multilinguality Pre-Quiz

JavaScript isn't enabled in your browser, so this file can't be opened. Enable and reload.

Email *

In an LSTM network, the two inputs at the forget gate represent

input at the particular time, and previous cell output

two inputs at the particular time

two previous cell inputs

weight matrix and bias

Clear selection

Addition of useful information to the cell state is done by

Output gate

Forget gate

Input gate and output gate

Input gate

Clear selection

In the Question Answering (RE) problem, the LSTM/GRU based layer act as:

Encoding layer

Embedding Layer

Attention

Bidirectional Attention

Clear selection

Context-to-Question (C2Q) Attention can be computed via:

Row-wise softmax of Similarity matrix to obtain attention distributions

Vector representation of Context

Vector representation of Question

Concatenated vector of Context and Questions

Clear selection

Question-to-Context(Q2C) Attention can be computed via:

Max of the corresponding row of the similarity matrix and then the softmax of the resulting vector

Vector representation of Question

Vector representation of Context

Concatenated vector of Context and Questions

Clear selection

Submit

Clear form

This content is neither created nor endorsed by Google. Report Abuse - Terms of Service - Privacy Policy