Paper Explained - Feedback Transformers: Addressing Some Limitations of Transformers with Feedback Memory

Feedback Transformers by @facebookai tackle one of the biggest shortcomings of Transformers They feed back information from higher to lower layers and use the model’s full capacity

Paper: [2002.09402] Addressing Some Limitations of Transformers with Feedback Memory

Abstract: Transformers have been successfully applied to sequential, auto-regressive tasks despite being feedforward networks. Unlike recurrent neural networks, Transformers use attention to capture temporal relations while processing input tokens in parallel. While this parallelization makes them computationally efficient, it restricts the model from fully exploiting the sequential nature of the input. The representation at a given layer can only access representations from lower layers, rather than the higher level representations already available. In this work, we propose the Feedback Transformer architecture that exposes all previous representations to all future representations, meaning the lowest representation of the current timestep is formed from the highest-level abstract representation of the past. We demonstrate on a variety of benchmarks in language modeling, machine translation, and reinforcement learning that the increased representation capacity can create small, shallow models with much stronger performance than comparable Transformers. Authors: Angela Fan, Thibaut Lavril, Edouard Grave, Armand Joulin, Sainbayar Sukhbaatar

3 Likes

Here to say I’m a big Yannic Kilcher fan + a fan of your aviators :smiley:

1 Like