Use basic and standard prompts to build a good basis.

Apply parameters to drive LLM responses.

Combine prompt components for powerful and focused interactions

Give clear instructions and avoid ambiguity

Describe exactly what you want to achieve with your prompt

Summarize, extract, and respond to information

Classify, Conversate, and Generate Code

Use logical reasoning in your prompts

Learn how to work with techniques that drive AI models, such as zero-shot, few-shot, and chain-of-thought prompting

Explore tools like Automatic Prompt Engineer (APE) for advanced prompts.

Generate data for analysis, training, or tests.

Get to know ChatGPT and its applications

Learn how to work with multi-turn conversations for in-depth interactions

Perform single-turn tasks for fast and efficient results

Prevent malicious actors from manipulating your AI via prompt injection

Protect sensitive information to prevent prompt leaking

Recognize techniques to bypass restrictions, such as prompt jailbreaking

Design clear and well-structured prompts to avoid misinterpretation

Test and improve prompts iteratively for consistent and reliable results

Analyze errors in output and adjust prompts to correct them

Optimise parameters such as temperature and tokens for better performance

Understand the context and make sure prompts match the task or target group

Formulate instructions clearly and concisely to work effectively

Think creatively and solve complex problems with innovative prompt designs

Test scenarios and vary prompts to find the most effective approach

Formulate prompts with specific and clear questions to get factual information

Verify the source and reliability of generated information

Test prompts for consistency by applying them repeatedly to the same task

Understand the limitations of the model and how it handles actual data

Recognize and minimize the risk of hallucinatory (incorrect) answers

Use external knowledge sources or data sources to verify output

Create prompts that direct the model to information processing rather than speculation

Be critical of output and manually validate key facts

Recognize possible biases in input and output

Test prompts with various scenarios and datasets

Minimize bias by designing neutral and inclusive prompts

Critically analyze output to identify unwanted patterns

Correct biases by modifying prompts or training data