Learning to reason with third-order tensor products

Imanol Schlag, Jürgen Schmidhuber

Research output: Chapter in Book/Report/Conference proceedingConference contribution

30 Scopus citations

Abstract

We combine Recurrent Neural Networks with Tensor Product Representations to learn combinatorial representations of sequential data. This improves symbolic interpretation and systematic generalisation. Our architecture is trained end-to-end through gradient descent on a variety of simple natural language reasoning tasks, significantly outperforming the latest state-of-the-art models in single-task and all-tasks settings. We also augment a subset of the data such that training and test data exhibit large systematic differences and show that our approach generalises better than the previous state-of-the-art.
Original languageEnglish (US)
Title of host publicationAdvances in Neural Information Processing Systems
PublisherNeural information processing systems foundation
Pages9981-9993
Number of pages13
StatePublished - Jan 1 2018
Externally publishedYes

Fingerprint

Dive into the research topics of 'Learning to reason with third-order tensor products'. Together they form a unique fingerprint.

Cite this