Learning the Context of Errors: Black-Box Online Adaptation of Time Series Foundation Models
For practitioners using closed-source TSFM APIs, this work enables effective online adaptation without model access, addressing a practical deployment bottleneck.
The paper addresses black-box online adaptation of time series foundation models (TSFMs) by learning the context of errors (input and output of the base model). The proposed method, ORCA, achieves improved forecasting performance across 5 TSFMs and 8 datasets, with ablation studies quantifying the impact of different adaptation hypotheses.
The rapid evolution of Time Series Foundation Models (TSFMs) has advanced zero-shot forecasting across diverse domains. Inspired by the current form of Large Language Models, future TSFMs may be offered as commercialized, closed-source API services. However, many existing online adaptation methods still rely on white-box access for parameter fine-tuning or gradient backpropagation. This paradigm mismatch raises a question: In black-box online adaptation for TSFMs, what should we learn? We answer this with an insight: the predictive errors of the base model are conditioned on both the input and output of the base model (i.e., the context of errors). To validate this insight, we propose ORCA (Online Residual Contextual Adaptation). We conduct extensive experiments across 5 state-of-the-art TSFMs and 8 datasets to demonstrate the effectiveness of our approach. Furthermore, through ablation studies, we quantitatively analyze the impact of different adapter learning hypotheses on the final adaptation performance in black-box online adaptation. Code available at https://github.com/Fifthky/ORCA.