Efficient Transformers with Dynamic Token Pooling

Citations Over TimeTop 10% of 2023 papers

Abstract

Transformers achieve unrivalled performance in modelling language, but remain inefficient in terms of memory and time complexity. A possible remedy is to reduce the sequence length in the intermediate layers by pooling fixed-length segments of tokens. Nevertheless, natural units of meaning, such as words or phrases, display varying sizes. To address this mismatch, we equip language models with a dynamic-pooling mechanism, which predicts segment boundaries in an autoregressive fashion. We compare several methods to infer boundaries, including end-to-end learning through stochastic re-parameterisation, supervised learning (based on segmentations from subword tokenizers or spikes in conditional entropy), as well as linguistically motivated boundaries. We perform character-level evaluation on texts from multiple datasets and morphologically diverse languages. The results demonstrate that dynamic pooling, which jointly segments and models language, is both faster and more accurate than vanilla Transformers and fixed-length pooling within the same computational budget.

Related Papers

• → A improved pooling method for convolutional neural networks(2024)119 cited
• Pooling in high-throughput drug screening.(2009)
• → A fully trainable network with RNN-based pooling(2019)22 cited
• Alpha-Pooling for Convolutional Neural Networks.(2018)
• → A Fully Trainable Network with RNN-based Pooling(2017)1 cited