“DataMUX: Data Multiplexing for Neural Networks”, 2023-01-13 (; backlinks):
We present data multiplexing (DataMUX)—a technique that enables deep neural networks to process multiple inputs simultaneously using a single compact representation and dramatically improves inference throughput
In this paper, we introduce data multiplexing (DataMUX), a technique that enables deep neural networks to process multiple inputs simultaneously using a single compact representation. DataMUX demonstrates that neural networks are capable of generating accurate predictions over mixtures of inputs, resulting in increased inference throughput with minimal extra memory requirements. Our approach uses two key components—(1) a multiplexing layer that performs a fixed linear transformation to each input before combining them to create a \“mixed\” representation of the same size as a single input, which is then processed by the base network, and (2) a demultiplexing layer that converts the base network’s output back into independent representations before producing predictions for each input.
We show the viability of DataMUX for different architectures (Transformers, and to a much lesser extent MLPs and CNNs) across 6 different tasks spanning sentence classification, named entity recognition and image classification.
For instance, DataMUX for Transformers can multiplex up to 20×/40× inputs, achieving up to 11×/18× increase in inference throughput with absolute performance drops of <2% and <4% respectively compared to a vanilla Transformer on MNLI, a natural language inference task.
We also provide a theoretical construction for multiplexing in self-attention networks and analyze the effect of various design elements in DataMUX.
[Keywords: neural networks, multiplexing, efficient inference]