Keywords: expire, long attention, memory, transformers
Abstract: Attention mechanisms have shown promising results in sequence modeling tasks that require long-term memory. Recent work has investigated mechanisms to reduce the computational cost of preserving and storing the memories. However, not all content in the past is equally important to remember. We propose Expire-Span, a method that learns to retain the most important information and expire the irrelevant information. This enables Transformers to scale to attend to tens of thousands of previous timesteps efficiently, as not all hidden states from previous timesteps are preserved. We demonstrate that Expire-Span can help models identify and retain critical information and show it can achieve state of the art results on long-context language modeling, reinforcement learning, and algorithmic tasks. Finally, we show that Expire-Span can scale to memories that are tens of thousands in size, which is helpful on incredibly long context tasks such as character-level PG-19 and a frame-by-frame moving objects task.
One-sentence Summary: Scale attention mechanisms in Transformers by learning what to forget from the past.
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Reviewed Version (pdf): https://openreview.net/references/pdf?id=cGL2wzQVP