The Power of Hard Attention Transformers on Data Sequences: A Formal Language Theoretic Perspective
Authors:
Pascal Bergsträßer,
Chris Köcher,
Anthony Widjaja Lin,
Georg Zetzsche
Abstract:
Formal language theory has recently been successfully employed to unravel the power of transformer encoders. This setting is primarily applicable in Natural Languange Processing (NLP), as a token embedding function (where a bounded number of tokens is admitted) is first applied before feeding the input to the transformer. On certain kinds of data (e.g. time series), we want our transformers to be…
▽ More
Formal language theory has recently been successfully employed to unravel the power of transformer encoders. This setting is primarily applicable in Natural Languange Processing (NLP), as a token embedding function (where a bounded number of tokens is admitted) is first applied before feeding the input to the transformer. On certain kinds of data (e.g. time series), we want our transformers to be able to handle \emph{arbitrary} input sequences of numbers (or tuples thereof) without \emph{a priori} limiting the values of these numbers. In this paper, we initiate the study of the expressive power of transformer encoders on sequences of data (i.e. tuples of numbers). Our results indicate an increase in expressive power of hard attention transformers over data sequences, in stark contrast to the case of strings. In particular, we prove that Unique Hard Attention Transformers (UHAT) over inputs as data sequences no longer lie within the circuit complexity class $AC^0$ (even without positional encodings), unlike the case of string inputs, but are still within the complexity class $TC^0$ (even with positional encodings). Over strings, UHAT without positional encodings capture only regular languages. In contrast, we show that over data sequences UHAT can capture non-regular properties. Finally, we show that UHAT capture languages definable in an extension of linear temporal logic with unary numeric predicates and arithmetics.
△ Less
Submitted 25 May, 2024;
originally announced May 2024.
