Story
arxiv_cs_lg · Aug 20, 2026 · paper
Source brief
Scale-Aware Pretraining of Time Series Foundation Models via Multi-Patch Token Alignment and Hybrid Masking
arxiv.orgAug 20, 2026
original source linked
In brief
Pretraining time series foundation models across heterogeneous datasets necessitates effective handling of varying sampling frequencies. Current methods either employ dataset-specific patch sizes and separate FFNs, le...
Feed lens
eval