EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

Three Tiers of Computation in Transformers and in Brain Architectures

Emma Graham, Richard Granger

ACM Transactions on Intelligent Systems and Technology · 28.02.2027 · 18 (1) · 1-28

DOI: 10.1145/3844612 ↗

Kurz gefasst

Fachartikel in der Zeitschrift ACM Transactions on Intelligent Systems and Technology, online erschienen am 29.10.2026. Verfasst von Emma Graham und 1 weiteren Autor:innen. Die Arbeit nennt 212 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: Office of Naval Research.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

Enabling AI systems to exhibit both fluent natural language and reliable logical reasoning is of great interest, yet the underlying machinery necessary for these capabilities remains poorly understood. Increases in abilities arise from many factors (scaling, architecture, training), yet the core required computational abilities for any particular task still elude understanding. We show that it is the transition between computational tiers in the automata hierarchy from the Theory of Computation, rather than scaling or particular algorithms, that determines a system’s capabilities. We provide a rigorous theoretical treatment and empirical evaluations of both language models and human subjects, culminating in a theoretically grounded pragmatic benchmarking method that can be used to ascertain the computational tier of an artificial or biological system. Specifically, we identify the tiers for the transition from prelinguistic to fluent natural language, and the tiers for the subsequent transition from articulate language use to reliable logical reasoning. The resulting principled analyses provide explanatory accounts of the abilities and shortfalls of a range of language models, and suggest actionable insights into the expansion of their logic and reasoning abilities.

Förderung

  • Office of Naval Research

Zitieren

APA

Graham, E., & Granger, R. (2027). Three Tiers of Computation in Transformers and in Brain Architectures. ACM Transactions on Intelligent Systems and Technology, 18(1), 1-28. https://doi.org/10.1145/3844612

Vancouver

Graham E, Granger R. Three Tiers of Computation in Transformers and in Brain Architectures. ACM Transactions on Intelligent Systems and Technology. 2027;18(1):1-28. doi:10.1145/3844612

BibTeX
@article{Graham2027,
  author = {Graham, Emma and Granger, Richard},
  title = {Three Tiers of Computation in Transformers and in Brain Architectures},
  journal = {ACM Transactions on Intelligent Systems and Technology},
  year = {2027},
  volume = {18},
  number = {1},
  pages = {1--28},
  doi = {10.1145/3844612}
}
Kostenlos starten →