​
Including or Excluding Public Knowledge

When you use Storytell, you can choose to include or exclude public knowledge from your SmartChats™. This allows you to utilize two distinct data sources to answer your queries:
  1. Your data: The assets you have uploaded into the Collection you are chatting with
  2. Public knowledge: The Large Language Model’s foundational training data
To incorporate public knowledge into your SmartChat™, ensure the Public Knowledge button is highlighted in your prompt bar. If you want to limit SmartChat™ to only use your uploaded Collection assets, select the appropriate button based on your needs, whether it’s solely from your Collections or a combination of your Collections and Public Knowledge. Storytell will tell you whether each answer includes or excludes public knowledge in the answer:

​
How Storytell Uses Public Knowledge

When you include public knowledge in your SmartChat™ thread, Storytell will prioritize the assets you’ve uploaded in the Collection while layering in additional context from the foundational training data the LLM we’ve selected to power your answer.
  • If the answer you are looking for is not contained in your uploaded assets, Storytell will look for it via the LLM’s foundational training data.
  • If the data contained in your uploaded assets conflicts with the LLM’s foundational training data, Storytell will prioritize your uploaded assets to provide you answers, but may layer in conflicting public knowledge
  • If you’re uncertain whether your answer is being overly influenced by public knowledge, you can deselect the Public Knowledge option in your prompt bar and choose your Collection to restrict the query to your uploaded data.