Learn Before
Using LLMs for Prompt Simplification
A straightforward method for simplifying prompts involves utilizing Large Language Models (LLMs) that have been previously fine-tuned or aligned for text simplification. This approach leverages the inherent capabilities of these models, treating prompt refinement as a direct text simplification task.
0
1
Tags
Ch.4 Alignment - Foundations of Large Language Models
Foundations of Large Language Models
Foundations of Large Language Models Course
Computing Sciences
Ch.3 Prompting - Foundations of Large Language Models
Related
Simplifying Task Descriptions by Deletion
Simplifying Task Descriptions by Paraphrasing
Using LLMs for Prompt Simplification
Example of a Detailed Task Description for a Healthcare and Finance QA Model
A developer is trying to make their instructions for a language model more efficient. Below is their original set of instructions. Which of the following revisions best simplifies the text while preserving all the essential constraints and the core intent of the original request?
Original Instructions: "I would be very grateful if you could please function as an expert travel planner. Your task is to create a comprehensive and detailed seven-day travel itinerary for a family of four, which includes two adults and two teenagers, who will be visiting the city of Paris for the very first time. It is important that the itinerary focuses on their interests in art, history, and food. Additionally, please make sure to incorporate some engaging and fun activities that would appeal specifically to the teenagers in the group."
Analyzing a Failed Prompt Simplification
Text Simplification as a Sequence-to-Sequence Task
Heuristic-Based Deletion for Prompt Simplification
Applying Prompt Simplification
Learn After
Constrained Prompting for Text Simplification
Refining a Complex Instruction
A developer has created a long and convoluted prompt for a text-to-image model, which is yielding inconsistent and low-quality results. They suspect the complexity of the prompt is the issue. Based on the principle of using a language model for simplification, which of the following actions represents the most effective and direct first step to address this problem?
Principle of Automated Prompt Refinement