You are building a question and answering system. You retrieve documents from a corpus given a question text. A large language model generates an answer based on these retrieved documents. What is this architecture for answering questions commonly called ?

Compute the interpolated precision as discussed in class for a recall of 0.5 from the given tuples. The data is presented in the following format : (Precision, Recall) 

(0.2, 0.1), (0.3, 0.6), (0.9, 0.7), (0.91, 0.2), (0.98, 0.73)

You are an engineer who has to build an information retrieval system for a doctor. The system can only retrieve a fixed number of documents at a time after which there is a long waiting period until a new set of documents can be retrieved. You want to minimize the time taken for the doctor to retrieve the most relevant document for their query. Which metric are you going to optimize to achieve this ?

Which of the following metric tends to fluctuate i.e. may increase or decrease as we go down the ranks of the set of retrieved documents in an information retrieval system ?

How will you manipulate the parameters k and b of BM25 to convert it to TF/IDF ?

Which combination of gates and operations in LSTM ensures that long-term dependencies

The words "bank," "river," and "finance" are projected into a 2D vector space. If "bank" is closer to "finance" than "river," what does this imply about the training corpus?

Given an LSTM network, if the forget gate outputs 0.1 and the cell state from the previous

During forward propagation in an RNN, the output at time step t is heavily influenced by

In an LSTM, assume the forget gate output Ft = 0.7, previous cell state Ct−1 = 0.9,