Motivated by recent advances in large language models for Natural Language Processing (NLP), we design a time-series foundation model for forecasting whose out-of-the-box zero-shot performance on a variety of public datasets comes close to the accuracy of state-of-the-art supervised forecasting models for each individual dataset. Our model is based on pretraining a patched-decoder style attention model on a large time-series corpus, and can work well across different forecasting history lengths, prediction lengths and temporal granularities.

https://huggingface.co/papers/2402.12659

simple dynamic webpages

If the fractional quantum Hall regime were a series of highways, these highways would have either two or four lanes. The flow of the two-flux or four-flux composite fermions, like automobiles in this two- to four-flux composite fermion traffic scenario, naturally explains the more than 90 fractional quantum Hall states that form in a large variety of host materials. Physicists at Purdue University have recently discovered, though, that fractional quantum Hall regimes are not limited to two-flux or four-flux and have discovered the existence of a new type of emergent particle, which they are calling six-flux composite fermion.

Our physical, 3D world consists of just two types of particles: bosons, which include light and the famous Higgs boson; and fermions—the protons, neutrons, and electrons that comprise all the "stuff," present company included.