Hello community, I am again here seeking some help. The confusion is about the fuzziness of TS. I was able to reproduce it in the popular https://linux-commits-search.typesense.org/ Query 1:

{
  "searches": [
    {
      "query_by": "subject,body",
      "query_by_weights": "1,1",
      "drop_tokens_threshold": 2,
      "typo_tokens_threshold": 2,
      "num_typos": 1,
      "sort_by": "",
      "highlight_full_fields": "subject,body",
      "collection": "commits",
      "q": "3d printing",
      "page": 1,
      "per_page": 5
    }
  ]
}

4 result! Query 2:

{
  "searches": [
    {
      "query_by": "subject,body",
      "query_by_weights": "1,1",
      "drop_tokens_threshold": 2,
      "typo_tokens_threshold": 2,
      "num_typos": 1,
      "sort_by": "",
      "highlight_full_fields": "subject,body",
      "collection": "commits",
      "q": "3d printing",
      "filter_by": "author_email_domain:=[`<http://amd.com|amd.com>`]",
      "page": 1,
      "per_page": 5
    }
  ]
}

38 results! I am facing this same issue in my current project as well. shouldn't the request without filter_by (Query 1) fetch more results? How to mitigate from this scenario? enabling exhaustive_search introduces severe performance hit. What should be a recommended, balanced and performant solution?