HiROv2 nails extremes: within 3% of X-SHiELD on 1-in-a-million wind, rain and pressure events
allen_ai · x · 2026-10-15
Ai2's thread continues with HiROv2's extremes performance: it matches X-SHiELD's distribution within 3% for 1-in-a-million rain, wind and pressure events, while capturing cyclone energy, realistic hurricanes and strong downslope winds. Demo link included, with side-by-side visuals (Santa Ana winds, tropical cyclones) comparing 100 km coarse output against 3 km downscaled views.
More from Research
- AlphaNIST: one RL policy solves all 20 NIST assembly tasks with 98.2% success in simulation — hbouammar · 2026-10-15
- Linear Depth Memory: matrix-valued memory tops PreLN and mHC on DepthBench across scales — SonglinYang4 · 2026-10-15
- Neural field planning for hybrid manufacturing cuts estimated build time 68.3% — jiqizhixin · 2026-10-15
- Tev-Gym: 1,170 dense-reward RL environments push long-horizon agentic training — RulinShao · 2026-10-15
- Claude systematically debunks the speculative EDAR-selenium hypothesis about East Asians — teortaxesTex · 2026-10-15
- Microsoft's AutoEnvScaling lets agents build their own RL environments, lifting Qwen3 Terminal-Bench from 41.1 to 53.3 — omarsar0 · 2026-10-15