Research map / Architectures and efficiency / Mixture of experts
VA-MoE: Variables-Adaptive Mixture of Experts for Incremental Weather Forecasting
Hao Chen, Han Tao, Guo Song, Jie Zhang, Yonghan Dong, Yunlong Yu, Lei Bai
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace plain MoE with VA-MoE (channel-adaptive experts) (ablation) | ViT+MoE · ERA5 1.5deg Z500 120h | RMSE: 356.02 -> 302.13 | Improved |
| replace ViT with VA-MoE (ablation) | ViT · ERA5 1.5deg Z500 120h | RMSE: 517.81 -> 302.13 | Improved |
| replace light MoE with VA-MoE (ablation) | ViT+MoE (light) · ERA5 1.5deg Z500 120h | RMSE: 405.73 -> 302.13 | Improved |
| replace ViT with VA-MoE (ablation) | ViT · ERA5 1.5deg T500 120h | RMSE: 2.40 -> 1.59 | Improved |
| incremental training of surface variables, 1979-2020 data, half iterations | VA-MoE · ERA5 0.25deg U10 120h | RMSE: 2.25 -> 2.33 | Hurt |
| incremental training of surface variables, 1979-2020 data, half iterations | VA-MoE · ERA5 0.25deg T2M 120h | RMSE: 1.42 -> 1.45 | Hurt |
| reduce incremental data to 2000-2020 and quarter of iterations | VA-MoE (IL) · ERA5 0.25deg MSL 120h | RMSE: 281.6 -> 304.7 | Hurt |
| reduce incremental data to 2000-2020 and quarter of iterations | VA-MoE (IL) · ERA5 0.25deg T2M 120h | RMSE: 1.45 -> 1.57 | Hurt |
| add incremental training of surface variables after upper-air training | VA-MoE · ERA5 0.25deg Z500 120h | RMSE: 295.52 -> 292.63 | Improved |
| add incremental training of surface variables after upper-air training | VA-MoE · ERA5 1.5deg Z500 120h | RMSE: 302.13 -> 301.41 | No clear effect |
| add incremental training of surface variables after upper-air training | VA-MoE · ERA5 0.25deg V500 72h | RMSE: 3.04 -> 3.51 | Hurt |
Closest papers by meaning
- EWMoE: An Effective Model for Global Weather Forecasting with Mixture-of-Experts (AAAI 2025 · similarity 0.87)
- Personalized Adapter for Large Meteorology Model on Devices: Towards Weather Foundation Models (NeurIPS 2024 · similarity 0.84)
- Time-MoE: Billion-Scale Time Series Foundation Models with Mixture of Experts (ICLR 2025 · similarity 0.83)
- Generalizing Weather Forecast to Fine-grained Temporal Scales via Physics-AI Hybrid Modeling (NeurIPS 2024 · similarity 0.83)
- Probabilistic Weather Forecasting with Hierarchical Graph Neural Networks (NeurIPS 2024 · similarity 0.81)
- Filtered not Mixed: Filtering-Based Online Gating for Mixture of Large Language Models (ICLR 2025 · similarity 0.81)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
