Reproduce “Cortex and subcortex play distinct roles over learning when cortical memory is limited” 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
Matthew Farrell, Taro Toyoizumi
The authors model how a limited memory for model-based learning affects adaptation when rewards move through a branching environment. They cap tracked transition edges and compare prioritizing reward-linked edges (MAXREWARD) with rootward edges (MAXREACH). In a depth-2, capacity-4 simulation, MAXREACH leads just after a reward switch, but MAXREWARD overtakes after about 25 episodes. The trade-off depends on environment structure.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.