LongEarth-R1, a vision-language model trained with sequence-aware supervision and spatiotemporal GRPO, sets new state-of-the-art results on all 12 tasks of the new LongEarth-Bench for long-horizon Earth observation reasoning.
AAAI , pages =
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
LongEarth-R1: Benchmarking and Aligning Vision-Language Models for Long-Horizon Earth Observation Reasoning
LongEarth-R1, a vision-language model trained with sequence-aware supervision and spatiotemporal GRPO, sets new state-of-the-art results on all 12 tasks of the new LongEarth-Bench for long-horizon Earth observation reasoning.