Nystr\"omformer: A Nystr\"om-Based Algorithm for Approximating Self-Attention

TL;DR

Nystr"omformer introduces a scalable self-attention approximation using the Nystr"om method, enabling efficient processing of long sequences in transformers while maintaining competitive performance on standard NLP benchmarks.

Contribution

It adapts the Nystr"om method to approximate self-attention with linear complexity, allowing transformers to handle longer sequences efficiently.

Findings

Performs comparably or better on GLUE and IMDB tasks.

Outperforms other efficient self-attention methods on Long Range Arena.

Achieves $O(n)$ complexity in self-attention approximation.

Abstract

Transformers have emerged as a powerful tool for a broad range of natural language processing tasks. A key component that drives the impressive performance of Transformers is the self-attention mechanism that encodes the influence or dependence of other tokens on each specific token. While beneficial, the quadratic complexity of self-attention on the input sequence length has limited its application to longer sequences -- a topic being actively studied in the community. To address this limitation, we propose Nystr\"{o}mformer -- a model that exhibits favorable scalability as a function of sequence length. Our idea is based on adapting the Nystr\"{o}m method to approximate standard self-attention with $O(n)$ complexity. The scalability of Nystr\"{o}mformer enables application to longer sequences with thousands of tokens. We perform evaluations on multiple downstream tasks on the GLUE benchmark and IMDB reviews with standard sequence length, and find that our Nystr\"{o}mformer performs comparably, or in a few cases, even slightly better, than standard self-attention. On longer sequence tasks in the Long Range Arena (LRA) benchmark, Nystr\"{o}mformer performs favorably relative to other efficient self-attention methods. Our code is available at https://github.com/mlpen/Nystromformer.