On the Ability and Limitations of Transformers to Recognize Formal Languages
This work addresses the problem of understanding Transformers' syntactic modeling capabilities for researchers in NLP and formal language theory, providing insights into their limitations compared to recurrent models.
The paper systematically studied Transformers' ability to model formal languages like counter and regular languages, finding they perform well on a subclass of counter languages but degrade on more complex regular languages, with performance correlating with a known complexity measure.
Transformers have supplanted recurrent models in a large number of NLP tasks. However, the differences in their abilities to model different syntactic properties remain largely unknown. Past works suggest that LSTMs generalize very well on regular languages and have close connections with counter languages. In this work, we systematically study the ability of Transformers to model such languages as well as the role of its individual components in doing so. We first provide a construction of Transformers for a subclass of counter languages, including well-studied languages such as n-ary Boolean Expressions, Dyck-1, and its generalizations. In experiments, we find that Transformers do well on this subclass, and their learned mechanism strongly correlates with our construction. Perhaps surprisingly, in contrast to LSTMs, Transformers do well only on a subset of regular languages with degrading performance as we make languages more complex according to a well-known measure of complexity. Our analysis also provides insights on the role of self-attention mechanism in modeling certain behaviors and the influence of positional encoding schemes on the learning and generalization abilities of the model.