Shared Memory

The Shared Memory Node is a special node designed to pass memory from one LLM (Large Language Model) node to another. Here’s how it works and what it does:

Purpose

• The Shared Memory node allows you to maintain conversational context or "memory" across multiple LLM nodes.

• It is especially useful if you want to share the conversation history, user input, or other relevant context between different AI models or across different parts of your workflow.

How it works in your flow

• You can choose how many of the past interactions it should pass. Be careful! Passing too many interactions may overwhelm the context window of the LLM you are passing to.

• Choose the LLM to pass from, you can choose from any of the LLMs in your current workflow. Any relevant context from previous steps or user inputs can be included in the LLM’s prompt, making the AI’s responses more coherent and context-aware.

• Reference the Shared Memory node in your LLM so the memory is passed in.

Typical Use Case

• If your workflow involves multiple turns of conversation or needs to remember previous user inputs, the Shared Memory node ensures that the LLM has access to this history, improving the quality and relevance of its responses.