Language Models - AI

What are the primary limitations of language models?

• Lack of data

• Complexity and variability of natural languages

• Insufficient computational power

• Lack of user input

What do language models provide for natural language tasks?

• Exact predictions

• Probabilistic predictions

• Definitive answers

• Fixed rules

The "bag-of-words" model treats a sentence as:

• A collection of individual words with context

• A sequence of words based on their order

• A grammatical structure

• A complex semantic unit

What type of language model is used for tasks like text classification?

• Bag-of-words model

• Character-level model

• Word representation model

In the bag-of-words model, how does it handle the order and context of words in a sentence?

• It captures both order and context effectively.

• It ignores the order and context of words.

• It assigns a fixed context to every word.

• It rearranges the words for better analysis

The bag-of-words model can be effective for:

• Identifying the order of words in a sentence

• Categorizing sentences into topics

• Capturing word interactions

• Parsing grammatical structures

What is the primary limitation of the bag-of-words model?

• Inability to assign probabilities to categories

• Inability to handle individual words

• Inability to recognize the importance of word sequences

• Inability to capture complex grammatical structures

Which type of N-gram model considers the probability of a word based on the previous word?

• Bigram model

• Trigram model

• 4-gram model

• Skip-gram model

What is one practical limitation of using extremely high values of 'n' in N-gram models?

• Improved accuracy

• Smaller model sizes

• Impractical model sizes

• Faster training times

Character-level models are useful for:

• Identifying languages and handling unknown words

• Analyzing semantic content of sentences

• Extracting information from images

• Creating complex grammatical structures