FlashKeepers

Machine Learning · College · College and University

Prompt Engineering, Embeddings and RAG

Prompt patterns, few-shot, chain of thought, embeddings, vector databases, chunking, retrieval augmented generation and agents.

About this set

40 basic cards for College Machine Learning. It opens with "What is a prompt pattern?" and "What does the role prompt pattern do?". Study it alongside the other College sets below.

40 cards · basic cards · AI-written, checked twice. Edit anything.

Study this set free Look inside first Get FlashKeepers for iPhone

Opens in your browser, no account. Get the iPhone and iPad app · Android app

What is a prompt pattern?
Show answerA reusable template for formulating prompts to achieve consistent, specific types of responses
What does the role prompt pattern do?
Show answerIt asks the LLM to adopt a specific persona or expertise level, such as 'You are a data scientist'
What is the context prompt pattern?
Show answerProviding background information or setting before the main task or question
What is the constraint prompt pattern?
Show answerSpecifying limitations or rules that the model must follow in its response
What is the output format prompt pattern?
Show answerExplicitly specifying how the response should be structured, such as JSON, a list, or a paragraph
What is the steps prompt pattern?
Show answerBreaking a complex task into a numbered sequence of smaller, manageable steps
What is the refinement prompt pattern?
Show answerAsking the model to improve, expand, or critique its own previous response
What is a meta-prompt?
Show answerA prompt that instructs the model how to handle or interpret subsequent prompts
What is few-shot learning?
Show answerProviding a small number of input-output examples in a prompt to teach the model desired behavior without fine-tuning
How does few-shot learning work?
Show answerThe model identifies patterns from the provided examples and applies those patterns to the new task
What is the difference between zero-shot and few-shot prompting?
Show answerZero-shot has no examples; few-shot includes a small number of input-output examples to demonstrate desired behavior
What is chain of thought (CoT) prompting?
Show answerA technique where the model is asked to show its reasoning step-by-step before providing the final answer
How does chain of thought improve accuracy?
Show answerIt forces systematic problem-solving, reducing errors and improving performance on complex reasoning tasks
What is zero-shot chain of thought?
Show answerAdding the phrase 'Let me think step by step' to a prompt, which encourages reasoning without explicit examples
What is the tree of thought technique?
Show answerAn approach that explores multiple reasoning paths in parallel and selects the most promising one to continue

25 more cards in the app

Study this set free