Reproduce “Spacetime Emergence in the Memory of a Learning Agent” 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

Philip A. LeMaitre, T. Rick Perche

The authors show an agent can recover spacetime structure from quantum-information feedback without receiving a metric or causal rules. The agent learns action preferences while moving a probe, and researchers fit light-cone directions from its averaged planner scores. In a flat, (1+1)-dimensional Minkowski setting, fitted null slopes are approximately ±0.999939, near ±1; this is post-processed readout, not a stored metric.

Summary by alphaXiv.

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