Channel Adapter for Time Series Foundation Models in Zero-Shot Multivariate Forecasting

Dongyuan Li, Renhe Jiang, Shun Zheng, Zheng Dong, Haotian Gao, Ying Zhang, Jiang Bian
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:67670-67698, 2026.

Abstract

Time Series Foundation Models (TSFMs) have achieved strong performance in univariate time series forecasting. However, most TSFMs rely on channel-independent pre-training that models each variable separately, limiting their ability to leverage inter-channel information that is crucial in real-world multivariate systems. Motivated by this limitation, we propose ChaTSFM, a lightweight plug-and-play channel adapter that allows frozen TSFMs to leverage multivariate correlations in a zero-shot setting. ChaTSFM first builds a budgeted pre-training dataset to cover diverse heterogeneous inter-channel dependency patterns. It then uses data-derived domain descriptors to learn a dataset-conditioned inter-channel similarity measure that reduces cross-domain metric distortion. Finally, it injects sparse inter-channel information via gated refinement, leveraging multivariate information without degrading intra-channel temporal dynamics. Extensive experiments on nine benchmarks validate the effectiveness of ChaTSFM, demonstrating consistent zero-shot improvements over four best-performing TSFMs while maintaining scalable deployment. Code is available at Code is available at https://github.com/Clearloveyuan/ChaTSFM.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-li26bu, title = {Channel Adapter for Time Series Foundation Models in Zero-Shot Multivariate Forecasting}, author = {Li, Dongyuan and Jiang, Renhe and Zheng, Shun and Dong, Zheng and Gao, Haotian and Zhang, Ying and Bian, Jiang}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {67670--67698}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/li26bu/li26bu.pdf}, url = {https://proceedings.mlr.press/v306/li26bu.html}, abstract = {Time Series Foundation Models (TSFMs) have achieved strong performance in univariate time series forecasting. However, most TSFMs rely on channel-independent pre-training that models each variable separately, limiting their ability to leverage inter-channel information that is crucial in real-world multivariate systems. Motivated by this limitation, we propose ChaTSFM, a lightweight plug-and-play channel adapter that allows frozen TSFMs to leverage multivariate correlations in a zero-shot setting. ChaTSFM first builds a budgeted pre-training dataset to cover diverse heterogeneous inter-channel dependency patterns. It then uses data-derived domain descriptors to learn a dataset-conditioned inter-channel similarity measure that reduces cross-domain metric distortion. Finally, it injects sparse inter-channel information via gated refinement, leveraging multivariate information without degrading intra-channel temporal dynamics. Extensive experiments on nine benchmarks validate the effectiveness of ChaTSFM, demonstrating consistent zero-shot improvements over four best-performing TSFMs while maintaining scalable deployment. Code is available at Code is available at https://github.com/Clearloveyuan/ChaTSFM.} }
Endnote
%0 Conference Paper %T Channel Adapter for Time Series Foundation Models in Zero-Shot Multivariate Forecasting %A Dongyuan Li %A Renhe Jiang %A Shun Zheng %A Zheng Dong %A Haotian Gao %A Ying Zhang %A Jiang Bian %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-li26bu %I PMLR %P 67670--67698 %U https://proceedings.mlr.press/v306/li26bu.html %V 306 %X Time Series Foundation Models (TSFMs) have achieved strong performance in univariate time series forecasting. However, most TSFMs rely on channel-independent pre-training that models each variable separately, limiting their ability to leverage inter-channel information that is crucial in real-world multivariate systems. Motivated by this limitation, we propose ChaTSFM, a lightweight plug-and-play channel adapter that allows frozen TSFMs to leverage multivariate correlations in a zero-shot setting. ChaTSFM first builds a budgeted pre-training dataset to cover diverse heterogeneous inter-channel dependency patterns. It then uses data-derived domain descriptors to learn a dataset-conditioned inter-channel similarity measure that reduces cross-domain metric distortion. Finally, it injects sparse inter-channel information via gated refinement, leveraging multivariate information without degrading intra-channel temporal dynamics. Extensive experiments on nine benchmarks validate the effectiveness of ChaTSFM, demonstrating consistent zero-shot improvements over four best-performing TSFMs while maintaining scalable deployment. Code is available at Code is available at https://github.com/Clearloveyuan/ChaTSFM.
APA
Li, D., Jiang, R., Zheng, S., Dong, Z., Gao, H., Zhang, Y. & Bian, J.. (2026). Channel Adapter for Time Series Foundation Models in Zero-Shot Multivariate Forecasting. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:67670-67698 Available from https://proceedings.mlr.press/v306/li26bu.html.

Related Material