Differentiable Top-k with Optimal Transport

By Yujia Xie, Hanjun Dai, Minshuo Chen, Bo Dai, Tuo Zhao, Hongyuan Zha, Wei Wei, Tomas Pfister. 2020

114 citations recorded by the source corpus.

Subfields: Optimization & Training

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Optimization & Training

1. Zhe Zhao
2. Lichan Hong
3. Ed H. Chi
4. Aakanksha Chowdhery
5. Hussein Hazimeh
6. Rahul Mazumder

Citations in the corpus

• DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning (2021, background): “Moreover, the stochastic k-subset selection framework in [25] (which encompasses several previous works) and the Top-k relaxation in [39] require solving an optimization subproblem to compute the gate output–each example will require solving a separate subproblem in the per-example gating setting, w”