FSPool: Learning Set Representations with Featurewise Sort Pooling

Yan Zhang, Jonathon Hare, Adam Prügel-Bennett

Keywords: set auto-encoder, set encoder, pooling

TL;DR: Sort in encoder and undo sorting in decoder to avoid responsibility problem in set auto-encoders

Abstract: Traditional set prediction models can struggle with simple datasets due to an issue we call the responsibility problem. We introduce a pooling method for sets of feature vectors based on sorting features across elements of the set. This can be used to construct a permutation-equivariant auto-encoder that avoids this responsibility problem. On a toy dataset of polygons and a set version of MNIST, we show that such an auto-encoder produces considerably better reconstructions and representations. Replacing the pooling function in existing set encoders with FSPool improves accuracy and convergence speed on a variety of datasets.

Code: https://github.com/Cyanogenoid/fspool

Community Implementations: [ 3 code implementations](https://www.catalyzex.com/paper/arxiv:1906.02795/code)