[ztratar]

Given the model performance is thus affected by a k-nearest neighbor, but those algorithms are proving not great for baseline vector search, how well will this actually work?

It seems mostly like a vertically integrated vector DB + existing LLM call, but correct me if I'm wrong. There are of course some performance gains with that, but the holy grail of "understanding" at unlimited length still seems unsolved.

[mrbungie]

Isn't the performance (as in the capacity of retrieval, not performance as compute/memory usage) of kNN mostly given by the quality of the vectors/embeddings themselves?

Most vector DBs use (at least) some kind of KNN anyways.