Reproduce “Chronosphere: Space-Time Tessellation of Local Climate Experts” 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
Daniel Cher, Eric Xing, Kexing Li, Brian Wei, Isaac Corley, Nathan Jacobs
Chronosphere represents locations across seasons by blending learned sites anchored to both place and time, while seasonal local experts add finer detail. The monthly model achieves 0.76 R² on snow-cover seasonal-window extrapolation, compared with 0.70 for CLIMPLICIT. The result comes from frozen encoders evaluated with cross-validated ridge probes, and shows an advantage on the seasonal transfer test, not future-climate prediction.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.