Metraipolis › archive › dc_579c3fd8ec83
Scale-Invariant Training for Time Series Foundation Models — arXiv:2610.07324
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.07324
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T11:20:32.583Z
- SHA-256 of the archived text
428c2da3d9c11f4cb62f6719b391476d3bad724d4d2423906884373d95d962d1- Stable URL of this page
https://metraipolis.com/documento/dc_579c3fd8ec83
How to cite
Scale-Invariant Training for Time Series Foundation Models — arXiv:2610.07324 — arXiv (en), https://arxiv.org/abs/2610.07324 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_579c3fd8ec83 (Metraipolis archive dc_579c3fd8ec83, sha256 428c2da3d9c11f4cb62f6719b391476d3bad724d4d2423906884373d95d962d1)
BibTeX
@misc{dc579c3fd8ec83,
title = {Scale-Invariant Training for Time Series Foundation Models — arXiv:2610.07324},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.07324},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_579c3fd8ec83; sha256 428c2da3d9c11f4cb62f6719b391476d3bad724d4d2423906884373d95d962d1},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.07324. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Time series foundation models (TSFMs) are trained on large collections of time series datasets that span various morphologies and domains. This setting exposes models to series whose scales -- typical magnitudes of their values -- can differ substantially. Affine scaling methods such as Reversible Instance Normalization (ReVIN) scale model inputs and reverse the transform before computing the loss. We show that this inversion multiplies each series' gradient by $b^p$ relative to loss on scaled targets, where $b$ is the scaling denominator (e.g., standard deviation) and $p$ is the loss degree. We call this scale-contaminated training (ScaleCon), because the scale of each series consequently becomes an importance weight, causing high-scale series to dominate training. For any scale-equivariant scaler and residual loss that is homogeneous of degree $p$, including MSE, MAE, and Quantile Loss, we prove that computing loss on scaled targets makes every mini-batch gradient and, consequently, the full optimization trajectory invariant to arbitrary independent rescaling of the training series, yielding scale-invariant training (ScaleIn). Notably, existing TSFMs use both objectives, with neither consistent reporting nor a common convention on how to compute training loss. We isolate the convergence disparity induced by ScaleCon and its correction under ScaleIn in controlled studies on synthetic and real data. In pretraining across four TSFM architectures, ScaleIn lowers MASE in all 24 architecture-benchmark comparisons, with average reductions across TSFMs of 18.8% on GIFT-Eval and 21.9% on the M-competitions. The gains extend to supervised neural forecasting, where it lowers MASE in 16 of 20 matched settings. Most existing time series forecasting pipelines can adopt ScaleIn with a one-line code change. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.07324 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-07T11:15:24.700Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).