Stable Attention Response for Reliable Precipitation Nowcasting
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866916008735801344 |
|---|---|
| author | Wen, Penghui Hu, Zexin Zhang, Sen Filippi, Patrick Zhu, Xiaogang Benter, Allen Bishop, Thomas Wang, Zhiyong Hu, Kun |
| author_facet | Wen, Penghui Hu, Zexin Zhang, Sen Filippi, Patrick Zhu, Xiaogang Benter, Allen Bishop, Thomas Wang, Zhiyong Hu, Kun |
| contents | Precipitation nowcasting remains challenging due to the highly localized, rapidly evolving, and heterogeneous nature of atmospheric dynamics. Although recent methods increasingly adopt attention-based architectures in both unimodal and multimodal settings, they mainly emphasize stronger representation learning and prediction capacity, while paying less attention to the stability of attention responses across samples. In this work, we show that cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability. Empirically, inaccurate forecasts are associated with larger attention-response energy variance across heads and layers. Theoretically, we show that cross-sample variability can propagate through self-attention, and enlarge a lower bound on prediction error. Based on this insight, we propose HARECast, a Head-wise Attention Response Energy-regulated framework for precipitation nowcasting. HARECast explicitly models head-wise attention-response energy and stabilizes it through a group-wise regularization objective that reduces cross-sample fluctuations. The proposed formulation is generic and applicable to both unimodal and multimodal nowcasting architectures. We instantiate HARECast in a standard forecasting pipeline with reconstruction branches and a diffusion-based predictor, and evaluate it on commonly used benchmarks--SEVIR and MeteoNet. Experimental results demonstrate that HARECast achieves state-of-the-art performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2605_13181 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Stable Attention Response for Reliable Precipitation Nowcasting Wen, Penghui Hu, Zexin Zhang, Sen Filippi, Patrick Zhu, Xiaogang Benter, Allen Bishop, Thomas Wang, Zhiyong Hu, Kun Machine Learning Artificial Intelligence Precipitation nowcasting remains challenging due to the highly localized, rapidly evolving, and heterogeneous nature of atmospheric dynamics. Although recent methods increasingly adopt attention-based architectures in both unimodal and multimodal settings, they mainly emphasize stronger representation learning and prediction capacity, while paying less attention to the stability of attention responses across samples. In this work, we show that cross-sample instability of attention-response energy is an important and previously underexplored source of forecasting unreliability. Empirically, inaccurate forecasts are associated with larger attention-response energy variance across heads and layers. Theoretically, we show that cross-sample variability can propagate through self-attention, and enlarge a lower bound on prediction error. Based on this insight, we propose HARECast, a Head-wise Attention Response Energy-regulated framework for precipitation nowcasting. HARECast explicitly models head-wise attention-response energy and stabilizes it through a group-wise regularization objective that reduces cross-sample fluctuations. The proposed formulation is generic and applicable to both unimodal and multimodal nowcasting architectures. We instantiate HARECast in a standard forecasting pipeline with reconstruction branches and a diffusion-based predictor, and evaluate it on commonly used benchmarks--SEVIR and MeteoNet. Experimental results demonstrate that HARECast achieves state-of-the-art performance. |
| title | Stable Attention Response for Reliable Precipitation Nowcasting |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2605.13181 |