Prompt Engineering Best Practices: LLM Output Validation & Evaluation

Validating Output from Instruction-Tuned LLMs

Checking outputs before showing them to users can be important for ensuring the quality, relevance, and safety of the responses provided to them or used in automation flows.

In this article, we will learn how to use the Moderation API by OpenAI to ensure safety and free of harassment output. Also, we will learn how to use additional prompts to the model to evaluate output quality before displaying them to the user to ensure the generated output follows the given instructions and is free of hallucinations.

This article is the ninth part of the ongoing series Prompt Engineering Best Practices:

1. Prompt Engineering Best Practices for Instruction-Tuned LLM [Part 1] 

2. Prompt Engineering Best Practices for Instruction-Tuned LLM [Part 2] 

3. Prompt Engineering for Instruction-Tuned LLM: Iterative Prompt Development 

4. Prompt Engineering for Instruction-Tuned LLM: Text Summarization 

5. Prompt Engineering for Instruction-Tuned LLM: Textual Inference & Sentiment Analysis 

6. Prompt Engineering for Instruction-Tuned LLM: Text Transforming & Translation

7. Prompt Engineering Best Practices: Text Expansion

8. Prompt Engineering Best Practices: Building Chatbots

9. Prompt Engineering Best Practices: Chain of Thought Reasoning

10. Prompt Engineering Best Practices: LLM Output Validation [You are here!]

Table of Contents:

1. Setting Up Working Environment & Getting Started

2. Checking Harmful Output 

3. Checking Instruction Following 

---

---