@inproceedings{xu2025unlocking,
author = {Xu, Chenhui and Yu, Fuxun and Bianco, Mike and Kovarskiy, Jacob and Tang, Raphael and Zhang, Qi and Xu, Zirui and Levine, William and Dubbs, Brandon and Liao, Heming and Burgess, C. and Bag, Suvam and Patravali, Jay and Kukal, Rupanjali and Figueroa, Mikael and Madhok, Rishi and Karianakis, Nikolaos and Xiong, Jinjun},
title = {Unlocking Zero-Shot Geospatial Reasoning via Indirect Rewards},
booktitle = {ICML 2026},
year = {2025},
month = {September},
abstract = {Training robust reasoning vision-language models (VLMs) in rare domains (such as geospatial) is fundamentally constrained by supervision scarcity. While raw geospatial imagery is abundant, the amount of task-direct supervision falls far behind that of common domains. In this work, we validate an important conclusion: indirect verifiable rewards, derived from seemingly unrelated metadata, are sufficient to induce sophisticated and generalizable geospatial reasoning across a wide range of downstream tasks (25+). We present Geo-R1 as one empirical instantiation of this paradigm. Rather than relying on limited task-specific annotations (i.e., direct rewards), Geo-R1 utilizes scalable, verifiable indirect proxy rewards based on cross-view alignment with metadata (geolocation information) to drive reinforcement learning at scale. Such indirect rewards successfully motivate the model to discover and internalize zero-shot geospatial reasoning across diverse tasks, achieving extraordinary zero-shot transfer on out-of-distribution benchmarks and even surpassing fully supervised specialists on certain benchmarks. These findings indicate that optimizing for indirect verifiable rewards may provide a scalable pathway to unlock generalized reasoning capabilities in rare domains with massive unlabeled data archives.},
url = {http://approjects.co.za/?big=en-us/research/publication/unlocking-zero-shot-geospatial-reasoning-via-indirect-rewards/},
}