Prompting Techniques | entAIngine documentation

📄️ Zero-Shot Prompting

Ask the model to perform a task without providing examples.

Guide the model with a few examples to shape its output.

Guide language models to reason step by step for complex tasks.

Generate multiple solutions and select the most consistent answer.

Use model-generated facts or context to improve responses.

Link multiple prompts to solve multi-step problems.

Explore multiple reasoning paths for better solutions.

Incorporate external knowledge into model responses.

Combine reasoning with tool use for complex tasks.

Use models to generate and refine prompts automatically for optimal performance.

Dynamically adjust prompts based on feedback, uncertainty, or changing context.

Guide the model's output with explicit cues, hints, and directional instructions.

Use code generation and execution to solve complex reasoning problems with high accuracy.

Integrate reasoning and action in an iterative loop for complex problem-solving.

Combine visual and textual reasoning for complex multimodal problems.

Structure information as graphs to enable relational reasoning and complex problem solving.