Reproduce “Causal Climate Emulation with Bayesian Filtering” 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
Sebastian Hickman, Ilija Trajkovic, Julia Kaltenborn, Francis Pelletier, Alex Archibald, Yaniv Gurwicz, Peer Nowack, David Rolnick and 1 more
PICABU emulates pre-industrial climate-model temperature by learning causal links among regional latent variables and filtering sampled long-term predictions against a climate statistic. In ten 50-year NorESM2 periods, its normalized GMST standard deviation and range were closer to ground truth than ViT’s, but remained higher; it did not match every statistic or establish real-world causal effects.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.