OSGR 框架提升 LLM 对地观测问答准确率
OSGR: Observation-Support Graph Reasoning for Earth Observation Question Answering
arXiv 上的新论文 OSGR，让 GPT-4o 在地理观测问答基准 Earth-Bench 上的准确率从 49.3% 拉到 63.7%，做时空检索推理的可以看看。
论文提出 OSGR，一种面向地球观测（EO）问答的两阶段框架，将检索问题分解为离线符号概念推导和在线 Relation-Conditioned Grounding。该框架利用离线 Observation-Support Graph 在概念层面推导目标到结果的支持路径，并通过时空兼容性核实例化观测项集合。在 Earth-Bench 和 DisasterM3-Reason 两个基准上，OSGR 将 GPT-4o 的准确率从 49.3% 提升至 63.7%，执行步骤比从 2.9 降至 2.5。
OSGR: Observation-Support Graph Reasoning for Earth Observation Question Answering
Answering geographic queries over Earth observation (EO) archives requires solving a joint retrieval-and-derivation problem under strict spatiotemporal constraints. Existing LLM agents fail to ground observation items or suffer from search-space explosion. We propose OSGR, a two-stage framework that factorizes this search into offline symbolic concept deduction and online Relation-Conditioned Grounding. OSGR leverages an offline Observation-Support Graph to deduce valid target-to-result support paths at the concept level, and applies spatiotemporal compatibility kernels to instantiate grounded observation-item sets. On Earth-Bench and DisasterM3-Reason, OSGR improves GPT-4o accuracy from 49.3\% to 63.7\% while reducing the execution-step ratio from 2.9 to 2.5.
