Knowledge Hub
Discover our library of insights and updates.
Spotlight Poster
Gridded Transformer Neural Processes for Spatio-Temporal Data
Our lead engineer Eric's spotlight paper: using gridded pseudo-token TNPs scale neural processes to large spatio-temporal datasets with efficient attention.
July 15, 2025
Article
A Decentralized Mixture-of-Experts Approach for State-of-the-Art Weather Forecasting
Using a robust Best Recent Performer (BRP) metric to identify consistently superior models, we benchmark the network against a dynamic state-of-the-art (SOTA) baseline.
July 17, 2025
Whitepaper
Zeus: Decentralized Climate Forecasting
The first whitepaper on Zeus: a decentralized climate forecasting platform that uses blockchain and machine learning to improve weather prediction accuracy and transparency.
February 7, 2025