Research

In natural language processing and computational linguistics, our research centers on a fundamental question: What are we doing when we model linguistic intelligence, and what can we learn from it? We approach this question by integrating perspectives from linguistics, psychology, and philosophy. Our current work focuses on the following three areas:

1. Evaluation of Linguistic Intelligence and Its Methodology

We develop datasets and methodologies for evaluating the behavior of language models. In particular, we aim to improve the reliability and validity of language model evaluation by incorporating frameworks from psychometrics.

Related Fields: Psychometrics, Education

2. Language Acquisition and Processing in Humans and Language Models

We investigate the principles underlying language acquisition and processing in both humans and language models through comparative analysis. A central goal is to understand how language models can achieve acquisition and processing that more closely resemble human behavior.

Related Fields: Language Acquisition, Psycholinguistics, Cognitive Science

3. Philosophy of Computational Linguistics

We examine what can be known through language model research, focusing on epistemological and ontological questions about intelligence and language understanding. We also consider the role of scientific models in the philosophy of science and their connections to constructs in psychology.

Related Fields: Philosophy of Science, Philosophy of Language/Linguistics, Epistemology