A research team wants to evaluate a new Large Language Model's ability to refine a given translation, specifically isolating this refinement skill from its ability to detect errors. They decide to use a simplified deliberate-then-generate approach. After providing the model with an original source sentence, what is the most appropriate next step in this specific methodology?
0
1
Tags
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models
Computing Sciences
Foundations of Large Language Models Course
Analysis in Bloom's Taxonomy
Cognitive Psychology
Psychology
Social Science
Empirical Science
Science
Related
Evaluating a Translation Improvement Strategy
Technique of Using Random Translation and Default Error Type
A research team wants to evaluate a new Large Language Model's ability to refine a given translation, specifically isolating this refinement skill from its ability to detect errors. They decide to use a simplified deliberate-then-generate approach. After providing the model with an original source sentence, what is the most appropriate next step in this specific methodology?
Critique of an LLM Evaluation Methodology