article

Transformers as Transducers
We characterize transformer sequence models as transducers and show some expressivity bounds for several variants.
Lena Strobl, Dana Angluin, David Chiang, Jon Rawski, Ashish Sabharwal
arXiv Preprint
Transformers as Transducers
The Problem-Ladenness of Theory
We give a pragmatic problem-centered account of theories’ epistemic virtues like coherence, depth, and parsimony.
Daniel Levenstein, Aniello De Santo, Saskia Heijnen, M. Narayan, Freek Oude Maatman, Jon Rawski, Cory wright
PDF OSF link
The Problem-Ladenness of Theory
Modern Language Models Refute Nothing
A short response rejecting the scientific contribution of language models as theories.
Jon Rawski, Lucie Baumont
Cite Preprint
Modern Language Models Refute Nothing
Computational Locality in Nonlinear Morphophonology
We present an automata-theoretic analysis of locality in nonlinear phonology and morphology.
Hossep Dolatian, Jon Rawski
Cite Review Draft
Computational Locality in Nonlinear Morphophonology
The Logical Nature of Phonology Across Speech and Sign
This article examines whether the computational properties of phonology hold across spoken and signed languages, using model theory and logical transductions.
Jon Rawski
Review Draft Invited Talk Slides Learning Slides
The Logical Nature of Phonology Across Speech and Sign