Data as of Jul 25, 2026 · Based on 2,718,867 AI responses across 9,511 prompts · See how Parse measures this
BRIGHTgithub.com · Low evidenceBRIGHT is a globally distributed, multimodal VHR dataset designed for AI-based disaster response across diverse weather conditions. It covers 14 disaster events in 23 regions, supporting supervised, transfer, and unsupervised learning methods.
Parse Score