A machine learning engineer is working with a large ensemble of language models. To generate a final prediction, they need to average the outputs over an intractably large space of possible input prompts. They decide to approximate this average by using a manageable, finite set of sample prompts. What is the fundamental trade-off inherent in this approximation strategy?
0
1
Tags
Ch.3 Prompting - 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
Estimating Model Performance Under Uncertainty
A machine learning engineer is working with a large ensemble of language models. To generate a final prediction, they need to average the outputs over an intractably large space of possible input prompts. They decide to approximate this average by using a manageable, finite set of sample prompts. What is the fundamental trade-off inherent in this approximation strategy?
Impact of Sample Size on Estimation Accuracy