If Attention Serves as a Cognitive Model of Human Memory Retrieval, What is the Plausible Memory Representation?

TL;DR

This paper explores whether Transformer Grammar's attention mechanism, which operates on syntactic structures, can serve as a cognitive model of human memory retrieval, showing it outperforms vanilla Transformers in predicting reading times.

Contribution

It introduces the use of syntactic-structure-based attention in Transformers as a cognitive model for human memory retrieval, emphasizing the role of syntactic structures as representational units.

Findings

TG's attention predicts reading times better than vanilla Transformers.

Both models contribute independently to predicting human sentence processing.

Human memory representations involve both syntactic structures and token sequences.

Abstract

Recent work in computational psycholinguistics has revealed intriguing parallels between attention mechanisms and human memory retrieval, focusing primarily on vanilla Transformers that operate on token-level representations. However, computational psycholinguistic research has also established that syntactic structures provide compelling explanations for human sentence processing that token-level factors cannot fully account for. In this paper, we investigate whether the attention mechanism of Transformer Grammar (TG), which uniquely operates on syntactic structures as representational units, can serve as a cognitive model of human memory retrieval, using Normalized Attention Entropy (NAE) as a linking hypothesis between models and humans. Our experiments demonstrate that TG's attention achieves superior predictive power for self-paced reading times compared to vanilla Transformer's, with further analyses revealing independent contributions from both models. These findings suggest that human sentence processing involves dual memory representations -- one based on syntactic structures and another on token sequences -- with attention serving as the general memory retrieval algorithm, while highlighting the importance of incorporating syntactic structures as representational units.