Learn Before
Example of a Prediction in Polarity Classification
As an example of the final step in the inference process for polarity classification, consider a model that outputs probabilities for 'positive', 'negative', and 'neutral' classes. If the probability for the 'positive' class is the highest among the three, the model's final prediction for the input text is 'positive'.
0
1
Tags
Ch.1 Pre-training - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Related
Example of Text for Polarity Classification
Output Formula for a Polarity Classification Model
Example of a Prediction in Polarity Classification
A model, which has been specialized for a text classification task, processes a new input and produces the following probability distribution over three possible classes:
{"Bug Report": 0.15, "Feature Request": 0.75, "General Inquiry": 0.10}. Based on this output, what is the model's final prediction?A language model has been specialized to classify customer support tickets into categories like 'Billing Issue', 'Technical Support', or 'Account Question'. Arrange the following steps in the correct sequence to describe how this model would process a new, unseen customer ticket to make a prediction.
Applying a Specialized Language Model
Learn After
A language model analyzes a customer review and produces the following probability distribution for its sentiment: 'positive': 0.15, 'negative': 0.78, 'neutral': 0.07. Based on this output, what is the model's final prediction for the review's sentiment?
Automated Feedback Categorization
Interpreting Model Output for Classification