Prof-K is a sample-then-filter top-k algorithm that reduces exact selection over N values to exact selection over a small candidate buffer with a distribution-free probabilistic guarantee.
The sparsely-gated mixture-of-experts layer.Outrageously large neural networks, 2:2, 2017
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
Prof-K: Probabilistic One-Pass Filtering for Efficient Top-k Selection
Prof-K is a sample-then-filter top-k algorithm that reduces exact selection over N values to exact selection over a small candidate buffer with a distribution-free probabilistic guarantee.