Prompt Engineering Techniques 💡
Explore advanced techniques to craft better prompts for Large Language Models.
Chain-of-Thought Prompting
Encourage models to explain their reasoning.
Tree of Thoughts
Use decision trees for reasoning paths.
Zero-shot Prompting
Directly generate answers without examples.
Few-shot Prompting
Provide few examples to guide the model.
Self-Consistency
Improve model confidence with consistent outputs.
Meta Prompting
Leverage multiple prompts to improve performance.
Prompt Chaining
Link prompts together for more complex tasks.
Generate Knowledge Prompting
Guide models to generate knowledge.
Retrieval Augmented Generation
Combine information retrieval with generation.
Automatic Prompt Engineer
Automate prompt creation and optimization.
Active-Prompt
Actively adjust prompts during the generation process.
Automatic Reasoning and Tool-use
Integrate reasoning with tool use.
Directional Stimulus Prompting
Direct model attention with stimulus.
Reflexion
Enable models to reflect on past outputs.
Multimodal CoT
Work with different types of inputs, including text and images.
Graph Prompting
Use graph structures to guide model responses.