Metric-Aware Hybrid Forecasting for the CTF4Science Lorenz Challenge 事件

Metric-Aware Hybrid Forecasting for the CTF4Science Lorenz Challenge arXiv:2606.04191v1 Announce Type: cross Abstract: We describe our approach to the CTF4Science Lorenz challenge, a benchmark that mixes short-horizon forecasting, long-time distribution matching, and trajectory reconstruction across nine task pairs. The key discovery is that no single model family dominated all metrics. Instead, we built a metric-aware hybrid system that assigned a different predictor to each metric family: (1)