GSR: Learning Structured Reasoning for Embodied Manipulation

Kewei Hu, Michael Zhang, Wei Ying, Tianhao Liu, Guoqiang Hao, Zimeng Li, Wanchan Yu, Jiajian Jing, Fangwen Chen, Hanwen Kang https://klmmotion.github.io/gsr.github.io/ · 2026

Read GSR: Learning Structured Reasoning for Embodied Manipulation by Kewei Hu, Michael Zhang, Wei Ying, Tianhao Liu, Guoqiang Hao, Zimeng Li, Wanchan Yu, Jiajian Jing, Fangwen Chen, Hanwen Kang https://klmmotion.github.io/gsr.github.io/ (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2602-01693.txt.

Description

Kewei Hu and colleagues give agents explicit world state reasoning, instead of burying task structure inside high dimensional latent representations. Their Grounded Scene graph Reasoning approach models world evolution as transitions over grounded scene graphs. It reasons stepwise over states and spatial relations rather than mapping perception to actions. That reasoning covers action preconditions, consequences and goals in a space. Training uses a dataset supervising understanding, action planning and goal interpretation. Tests across RLBench, LIBERO, a GSR benchmark and robots show better zero shot generalization and long completion than baselines.

Catalog record

CREATORKewei Hu, Michael Zhang, Wei Ying, Tianhao Liu, Guoqiang Hao, Zimeng Li, Wanchan Yu, Jiajian Jing, Fangwen Chen, Hanwen Kang https://klmmotion.github.io/gsr.github.io/
YEAR2026
COLLECTIONPAPERS
FORMATTXT
FILE SIZE63598
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2602.01693

Browse the full library · About the Tower of Babel