TempoPFN: Synthetic Time Series Generators and Augmentations for DeltaProduct-Based Linear RNN Foundation Models

·2026.10.19 13:12

Key point

The NeurIPS 2026 paper provides a library of synthetic time series generators and augmentations for training a DeltaProduct based linear RNN.

TempoPFN is a NeurIPS 2026 paper that presents a library of synthetic time series generators and augmentations for training time series foundation models, specifically a DeltaProduct-based linear RNN.
The synthetic data has already been used in Toto-2 and TS-ICL.
Resources include arXiv 2510.25502, GitHub automl/TempoPFN, and Hugging Face AutoML-org/TempoPFN.

This summary was generated automatically by AI. Check the original for the author's claims and context. Copyright belongs to the original author.

Our guide explains how the AI works. Report summary errors, attribution issues, or removal requests via Contact.