Anyway to pass minRelevanceScore to SearchEmbeddingAsync of IVectorStore?

[hanhsia]

Since IMemoryStore is legacy, I want to migrate my IMemoryStore code to IVectorStore, But I can't find a way to pass minRelevanceScore to SearchEmbeddingAsync, just like what SearchAsync of IMemoryStore did. Is there any technique to make the search function return results with a score above a certain level?
The previous IMemoryStore code:
await foreach (MemoryQueryResult memory in textMemory.SearchAsync(index, input, limit: resultLimit, minRelevanceScore: relevanceThreshold))
{
...
}
Now IVectorStore code:
await foreach (var memory in collection.SearchEmbeddingAsync(searchVector, top: resultLimit))
{
...
}

[markwallace-microsoft]

@westey-m can you help here?

[westey-m]

@hanhsia, unfortunately this is not currently supported. We do have an item on our backlog for this, but is is non-trivial from a usage perspective since the threshold will vary by both the database and distance function chosen, making it very hard to use in a database agnostic way. Note that we do return scores in the search results so secondary filtering is certainly still possible.

See #9566

CC @roji