A language model is being trained to determine if two sentences are consecutive. For 'positive' examples, it is given two sentences that appear one after the other in a book. For 'negative' examples, the first sentence is from a book about astrophysics, and the second is always from a children's fairy tale. What is the most significant risk associated with this training design?
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
Analysis in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Analysis of a Language Model's Training Objective
A language model is being trained to determine if two sentences are consecutive. For 'positive' examples, it is given two sentences that appear one after the other in a book. For 'negative' examples, the first sentence is from a book about astrophysics, and the second is always from a children's fairy tale. What is the most significant risk associated with this training design?
Evaluating a Model's Training Task