Improving differentiable neural computers through memory masking, de-allocation, and link distribution sharpness control

Róbert Csordás, Jürgen Schmidhuber

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

13 Scopus citations

Abstract

The Differentiable Neural Computer (DNC) can learn algorithmic and question answering tasks. An analysis of its internal activation patterns reveals three problems: Most importantly, the lack of key-value separation makes the address distribution resulting from content-based look-up noisy and flat, since the value influences the score calculation, although only the key should. Second, DNC's deallocation of memory results in aliasing, which is a problem for content-based look-up. Thirdly, chaining memory reads with the temporal linkage matrix exponentially degrades the quality of the address distribution. Our proposed fixes of these problems yield improved performance on arithmetic tasks, and also improve the mean error rate on the bAbI question answering dataset by 43%.
Original languageEnglish (US)
Title of host publication7th International Conference on Learning Representations, ICLR 2019
PublisherInternational Conference on Learning Representations, ICLR
StatePublished - Jan 1 2019
Externally publishedYes

Fingerprint

Dive into the research topics of 'Improving differentiable neural computers through memory masking, de-allocation, and link distribution sharpness control'. Together they form a unique fingerprint.

Cite this