Metraipolis › archive › dc_b15ee883f306
Does Joint-Embedding Predictive Architecture Pretraining Help Time Series Forecasting? — arXiv:2609.31680
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.31680
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T20:59:23.120Z
- SHA-256 of the archived text
41349e26e0c7db6e2c67393b4c6c9537b2df7f2a96f3257f37a8cf80cc3100b4- Stable URL of this page
https://metraipolis.com/documento/dc_b15ee883f306
How to cite
Does Joint-Embedding Predictive Architecture Pretraining Help Time Series Forecasting? — arXiv:2609.31680 — arXiv (en), https://arxiv.org/abs/2609.31680 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_b15ee883f306 (Metraipolis archive dc_b15ee883f306, sha256 41349e26e0c7db6e2c67393b4c6c9537b2df7f2a96f3257f37a8cf80cc3100b4)
BibTeX
@misc{dcb15ee883f306,
title = {Does Joint-Embedding Predictive Architecture Pretraining Help Time Series Forecasting? — arXiv:2609.31680},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.31680},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_b15ee883f306; sha256 41349e26e0c7db6e2c67393b4c6c9537b2df7f2a96f3257f37a8cf80cc3100b4},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.31680. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Joint-embedding predictive architectures (JEPA) have emerged as a promising self-supervised pretraining paradigm for time series, learning representations by predicting target embeddings in latent space rather than reconstructing raw signals. Yet evidence on their benefits remains mixed, and most studies test only a single backbone or a narrow set of architectures, leaving unclear whether JEPA pretraining is a reliable improvement or one that depends heavily on the downstream model. We address this gap through a large scale evaluation of one JEPA instantiation across nine backbones and eleven benchmarks spanning temporal and spatio-temporal forecasting, the most extensive cross architecture assessment of JEPA for time series to date. We find that the benefit of this instantiation varies sharply across backbones, producing consistent gains for some architectures and consistent degradation for others, even on the same dataset. This pattern holds across both task families, indicating the variability is a general property of this instantiation rather than a dataset specific artifact worth accounting for when choosing a backbone in practice. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI). Categorie arXiv: Machine Learning (cs.LG) ; Artificial Intelligence (cs.AI); Machine Learning (stat.ML). Pagina del lavoro: https://arxiv.org/abs/2609.31680 — 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 2026-10-07T20:29:02.027Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).