# Successive Halving Top-k Operator

arXiv: https://arxiv.org/abs/2010.15552 · HTML: https://mlsanity.com/paper/2010.15552/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 8 Oct 2020.

## Built on by (most cited first, 2 of 2)

- [Differentiable Top-k Classification Learning](https://mlsanity.com/paper/2206.07290/) (2022)
- [Learning to Summarize Videos by Contrasting Clips](https://mlsanity.com/paper/2301.05213/) (2023)

## Abstract (arXiv)

We propose a differentiable successive halving method of relaxing the top-k operator, rendering gradient-based optimization possible. The need to perform softmax iteratively on the entire vector of scores is avoided by using a tournament-style selection. As a result, a much better approximation of top-k with lower computational cost is achieved compared to the previous approach.