The results are ranked according to Limo's ranking principles (based on relevance and availability, among other things). However, the Large Language Model uses a different sorting principle based on a content comparison between the abstracts of the sources and the keywords in your search query.

So when you click on ‘View all results in Limo’, the sources at the top of the results list are usually different from the five listed in the overview. In addition to the different sorting algorithm, there are other reasons for this:

  • The search now covers the entire collection, including your library's local collections and other sources that are not accessible to the Limo AI Assistant.
  • By default, you are searching the full text of a large number of sources, while the AI assistant only searches the abstracts.