Reproduce “The Min-Max theory of language behavior” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Hiram Ring
The authors test how noun and verb character length and information weighting affect word order in a scratch-trained transformer using synthetic intransitive sentences. At α=5, a one-character noun with a five-character verb yields 50.69% SV bias, versus 49.58% when both are one character. These small model preferences do not establish human processing or historical change.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.