Iterate: LLM Knowledgebase
To provide the user a way to ask questions that can be answered by a LLM using a predefined knowledgebase
Last updated
DialoX Manual
To provide the user a way to ask questions that can be answered by a LLM using a predefined knowledgebase
Last updated
The LLM Knowledgebase node is used in the same way the default Iterate or Iterate: LLM node is used. This means, it is most often used as a form of main menu of your Smart App: "What can I do for you today?".
Large language models are great at understanding user questions and provide sensible answers based on the information stored in the common knowledge of the language model. The Iterate: LLM node solely relies on a predefined but limited prompt instruction to answer the user. This node however will also use a Knowledgebase on top of this prompt instruction. This knowledgebase can be configured in the AI section.
A Large Language Model (LLM) is a type of artificial intelligence (AI) model designed to understand and generate human-like text based on the input it receives. These models are trained on massive datasets of text from books, articles, websites, and other written content, enabling them to process and generate coherent language.
This node uses the ChatGPT models from Open AI together with the Assistants API, hosted on the Microsoft Azure platform to answer the users of your Smart app.
A knowledge base is a centralized repository for information, designed to store, organize, and manage knowledge in a way that makes it easily accessible and usable for a specific purpose. Knowledge bases are commonly used by organizations, apps, and systems to provide self-service support, streamline workflows, and enhance decision-making.
As explained in the introduction, the Iterate: LLM Knowledgebase node is working the same way as the Iterate: LLM node, adding the ability to use a predefined knowledgebase as its knowledge. Also, the Iterate: LLM Knowledgebase works as a menu. It either responds due to one of the labelled links or dialog triggers are triggered or otherwise respond from the LLM.
In the above example, the Iterate: LLM knowledgebase will match the intents associated with the two labelled links "Talk to a human" and "Done". When the user doesn't appear to be asking for one of these intents, it will use the LLM to generate an answer.
The LLM can use a specific instruction from you (as the Smart App designer) to guide the generation of the answer. This instruction is given in the form of a "Prompt". The prompt can be configured in the AI section of the platform. Besides instructions, the prompt will also contain content that the LLM can use on top of its own dataset to generate answers, like information regarding your address, opening hours, parking instructions, etc. See the Prompts page for more information on this topic.
Question
The initial question
Repeat (1)
The text to use when re-prompting the user after a turn
optional
Repeat (2)
The next text to use when re-prompting the user after a turn. The system will keep on iterating this step the third and all other turns
Prompt
The prompt which contains the instructions and content the LLM can use when generating a response
The knowledge base limitations are described in the knowledge base page in the AI section.
