Learn Before
Expansion in Prompt Search
In prompt optimization, expansion is a key operation used to explore different states in the search space. It can be defined as a function , where a set of new candidate prompts is generated from the existing pool using a model . When is a Large Language Model, it is instructed to generate new, relevant prompts based on the ones currently in , after which is replaced with . This process can be repeated iteratively alongside evaluation and pruning.
0
1
Tags
Ch.3 Prompting - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Related
Iterative LLM-Based Prompt Search
Expansion in Prompt Search
Applying Classic Optimization Techniques to Prompt Optimization
A team is developing a system to automatically find the best prompt for summarizing legal documents. Their process is as follows:
- They create a large, diverse list of 100 potential prompts.
- They use a small, representative dataset to calculate an accuracy score for each of the 100 prompts.
- They select the prompt with the highest accuracy score from the initial list and the process concludes.
Which critical element of an effective search strategy is missing from their approach?
Evaluating Prompt Search Strategies
Critique of a Prompt Finding Method
Learn After
Example of a Prompt for LLM-based Prompt Expansion
Iterative LLM-Based Prompt Search
A team is using an automated process to discover a high-performing prompt for a text summarization task. They begin with an initial set of prompts:
C = {'Summarize the following text for me.', 'Give me the main points of this article.'}. They then apply a single 'expansion' operation, which uses a large language model to generate a new set of candidate prompts based on the ones inC. Which of the following best represents a plausible output set from this single expansion operation?The Role of Expansion in Prompt Diversity
Expansion Function in Search Algorithms
Troubleshooting a Prompt Optimization Process
Prompt Expansion via Edit Operations
Prompt Expansion via Feedback
Prompt Paraphrasing