Timezone: »

Learning to Transduce with Unbounded Memory
Edward Grefenstette · Karl Moritz Hermann · Mustafa Suleyman · Phil Blunsom

Mon Dec 07 04:00 PM -- 08:59 PM (PST) @ 210 C #16

Recently, strong results have been demonstrated by Deep Recurrent Neural Networks on natural language transduction problems. In this paper we explore the representational power of these models using synthetic grammars designed to exhibit phenomena similar to those found in real transduction problems such as machine translation. These experiments lead us to propose new memory-based recurrent networks that implement continuously differentiable analogues of traditional data structures such as Stacks, Queues, and DeQues. We show that these architectures exhibit superior generalisation performance to Deep RNNs and are often able to learn the underlying generating algorithms in our transduction experiments.

Author Information

Edward Grefenstette (Google DeepMind)
Karl Moritz Hermann (Google DeepMind)
Mustafa Suleyman (Google DeepMind)
Phil Blunsom (Google DeepMind)

More from the Same Authors