Research map / Label-efficient and robust learning / Domain adaptation and generalization
Meta-Learning Dynamics Forecasting Using Task Inference
Rui Wang, Robin Walters, Rose Yu
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 |
|---|---|---|---|
| add magnitude loss to encoder training (vs Sup_c+time_inv) (ablation) | DyAd · Turbulent Flows future | RMSE: 0.51 -> 0.42 | Improved |
| add time-invariance and magnitude losses to encoder training (vs Sup_c only) (ablation) | DyAd · Turbulent Flows future | RMSE: 0.53 -> 0.42 | Improved |
| add time-invariance loss on top of Sup_c (ablation) | DyAd · Turbulent Flows future | RMSE: 0.53 -> 0.51 | Improved |
| use AdaIN instead of concatenating z_c to input (ConI) (ablation) | DyAd · Turbulent Flows future | RMSE: 1.09 -> 0.42 | Improved |
| use AdaIN instead of KGen kernels (ablation) | DyAd · Turbulent Flows future | RMSE: 0.78 -> 0.42 | Improved |
| use AdaIN instead of concatenating z_c to hidden states (ConH) (ablation) | DyAd · Turbulent Flows future | RMSE: 0.75 -> 0.42 | Improved |
| use AdaIN instead of adding z_c to hidden states (Sum) (ablation) | DyAd · Turbulent Flows future | RMSE: 0.84 -> 0.42 | Improved |
| use AdaIN instead of conv biases (Bias) (ablation) | DyAd · Turbulent Flows future | RMSE: 0.84 -> 0.42 | Improved |
| use AdaIN instead of ConI on domain test (ablation) | DyAd · Turbulent Flows domain | RMSE: 0.85 -> 0.51 | Improved |
| add encoder (vs No_enc, forecaster with AdaIN/AdaPad only) (ablation) | DyAd · Turbulent Flows future | RMSE | Improved |
| add AdaPad on top of AdaIN (ablation) | DyAd · Turbulent Flows future | RMSE | Improved |
| feed encoder correct weak supervision c instead of random c (Wrong_enc) (ablation) | DyAd · Turbulent Flows future | RMSE | Improved |
Closest papers by meaning
- Time Series Forecasting Through the Lens of Dynamics (ICML 2026 · similarity 0.87)
- Meta-Learning Framework with Applications to Zero-Shot Time-Series Forecasting (AAAI 2021 · similarity 0.87)
- Towards a General Time Series Forecasting Model with Unified Representation and Adaptive Transfer (ICML 2025 · similarity 0.86)
- Adaptive Meta-Learning Probabilistic Inference Framework for Long Sequence Prediction (AAAI 2024 · similarity 0.85)
- Meta-DT: Offline Meta-RL as Conditional Sequence Modeling with World Model Disentanglement (NeurIPS 2024 · similarity 0.84)
- TimePerceiver: An Encoder-Decoder Framework for Generalized Time-Series Forecasting (NeurIPS 2025 · similarity 0.84)
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.
