Y
Hacker News
new
|
ask
|
show
|
jobs
by
andrewlu0
1168 days ago
Will check this out - it seems like its doing two separate searches where you specify two separate top K values. Curious what the trade offs are between this approach and weighting sparse/dense vectors