Nlp Highlights

129 - Transformers and Hierarchical Structure, with Shunyu Yao

Informações:

Synopsis

In this episode, we talk to Shunyu Yao about recent insights into how transformers can represent hierarchical structure in language. Bounded-depth hierarchical structure is thought to be a key feature of natural languages, motivating Shunyu and his coauthors to show that transformers can efficiently represent bounded-depth Dyck languages, which can be thought of as a formal model of the structure of natural languages. We went on to discuss some of the intuitive ideas that emerge from the proofs, connections to RNNs, and insights about positional encodings that may have practical implications. More broadly, we also touched on the role of formal languages and other theoretical tools in modern NLP. Papers discussed in this episode: - Self-Attention Networks Can Process Bounded Hierarchical Languages (https://arxiv.org/abs/2105.11115) - Theoretical Limitations of Self-Attention in Neural Sequence Models (https://arxiv.org/abs/1906.06755) - RNNs can generate bounded hierarchical languages with optimal memory (ht