![[2602.16188] Deep TPC: Temporal-Prior Conditioning for Time Series Forecasting](https://arxiv.org/static/browse/0.3.4/images/arxiv-logo-fb.png){kind=logo}
[2602.16188] Deep TPC: Temporal-Prior Conditioning for Time Series Forecasting
Summary
The paper introduces Temporal-Prior Conditioning (TPC) for time series forecasting, enhancing temporal reasoning by integrating time as a key modality throughout the model layers, leading to superior performance in long-term forecasting tasks.
Why It Matters
As time series forecasting becomes increasingly vital across various sectors, TPC's innovative approach addresses limitations in current models, potentially transforming how temporal data is processed and improving forecasting accuracy. This advancement could benefit industries reliant on precise predictions.
Key Takeaways
- TPC treats time as a primary modality, enhancing model depth.
- The method outperforms traditional fine-tuning and shallow conditioning.
- It maintains a low parameter budget while improving forecasting accuracy.
- Cross-attention modules are trained specifically to disentangle signals.
- TPC achieves state-of-the-art results across diverse datasets.
Computer Science > Machine Learning arXiv:2602.16188 (cs) [Submitted on 18 Feb 2026] Title:Deep TPC: Temporal-Prior Conditioning for Time Series Forecasting Authors:Filippos Bellos, NaveenJohn Premkumar, Yannis Avrithis, Nam H. Nguyen, Jason J. Corso View a PDF of the paper titled Deep TPC: Temporal-Prior Conditioning for Time Series Forecasting, by Filippos Bellos and 4 other authors View PDF HTML (experimental) Abstract:LLM-for-time series (TS) methods typically treat time shallowly, injecting positional or prompt-based cues once at the input of a largely frozen decoder, which limits temporal reasoning as this information degrades through the layers. We introduce Temporal-Prior Conditioning (TPC), which elevates time to a first-class modality that conditions the model at multiple depths. TPC attaches a small set of learnable time series tokens to the patch stream; at selected layers these tokens cross-attend to temporal embeddings derived from compact, human-readable temporal descriptors encoded by the same frozen LLM, then feed temporal context back via self-attention. This disentangles time series signal and temporal information while maintaining a low parameter budget. We show that by training only the cross-attention modules and explicitly disentangling time series signal and temporal information, TPC consistently outperforms both full fine-tuning and shallow conditioning strategies, achieving state-of-the-art performance in long-term forecasting across diverse datas...
