Unverified70% confidenceFactExact time
The Prithvi model was jointly developed by IBM and NASA, trained primarily on Harmonized Landsat-Sentinel (HLS) data, and demonstrates excellent representational capability for medium-resolution multispectral land surface observations.
2
Sources
70%
Confidence
Medium-term (~90 days)
Relevance
8/8/2026
First Seen
Valid until: 11/6/2026
Sources
嵌入向量时代：数据科学从特征工程到表征选择的范式转移
Related Claims
UnverifiedUnitree Robotics open-sourced the visual-language-action large model Unifor L-MVLA-0, built on Qwen2.5VL 7B, with performance comparable to Gemini Robotics ER.56% similarUnverifiedH100/H200是英伟达专为大模型训练与推理设计的数据中心级GPU，采用NVLink高速互联技术使数千张卡协同工作54% similarUnverifiedF´直接应用于NASA机智号火星直升机（Ingenuity）的飞行控制软件54% similarUnverifiedU1 Pro targets design scenarios, supports long-cycle generation review loops, and can achieve output resolutions of up to 8K.54% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/711031API
curl https://kongchang.com/api/v1/knowledge/claims/711031MCP
get_claim(id=711031)