Reproduce “Building, Reusing, and Generalizing Abstract Representations from Concrete Sequences” 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

Shuchen Wu, Mirko Thalmann, Peter Dayan, Zeynep Akata, Eric Schulz

HVM learns chunks from observed sequences, then identifies chunks that occur between the same neighboring chunks as alternatives in a variable. This extends the Hierarchical Chunking Model, which learns chunks without variables. In a color-sequence experiment, HVM likelihood correlated with human recall time on transfer (R = 0.7, p ≤ 0.001), an association limited to this task, not proof of general human-like abstraction.

Summary by alphaXiv.

Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.