Graph-Operator World Models (GraphOp-WM) are a structured approach to learning world models that generalize across varying morphology parameters in continuous control tasks. The model represents robot bodies and their kinematic relationships as an attributed graph, decomposing each transition into a morphology‑independent local dynamics basis and a morphology‑conditioned structured operator. This operator blends node‑local modulation, kinematic‑tree coupling, and a low‑rank global correction, while architectural design choices encourage the operator to capture static morphology dependence. The framework supports reward, value, and TD‑MPC‑style planning through graph‑level readout and edge‑wise action representations, and is evaluated on controlled MuJoCo parameter splits involving interpolation, extrapolation, and held‑out compositions of link geometry, mass, damping, and actuation in Hopper, Walker2d, and HalfCheetah.
By Xu Yang, Yiqin Yang, Qianchuan Zhao
arXiv:2604. 08780v2 Announce Type: replace-cross Abstract: World models promise a paradigm shift in robotics, where an agent learns the physics of its environment once and then acquires behaviors efficiently.
By Mohamad H. Danesh, Chenhao Li, Amin Abyaneh, Anas Houssaini, Kirsty Ellis, Glen Berseth, Marco Hutter, Hsiu-Chin Lin
arXiv:2608. 01506v1 Announce Type: cross Abstract: Humans readily adapt their movements as their bodies change through aging, injury, or load carrying, but learning-based robot policies often break when hardware properties shift.
By Dichen Li, Bo Ai, Nico Bohlinger, Jan Peters, Hao Su, Henrik I. Christensen
RecMorph introduces a topology‑guided spatial recurrent architecture for generalized morphology control, converting a kinematic tree into a sequence that enables joint cross‑limb communication and representation transformation. The design incorporates residual preservation, RMS normalization, and input‑dependent channel modulation to stabilize repeated spatial transformations, achieving linear token complexity. Across five UNIMAL tasks and a four‑platform quadruped setting, RecMorph outperforms existing controllers in training performance, inference throughput, and generalization to unseen bodies with up to 30 limbs, while also demonstrating robust real‑world performance on Go1/Go2 trials.
By Quanrui Rao, Yong Liu, Xueming Xiao, Yingbo Luo, Kun Wu, Zhenyu Xu, Meibao Yao
arXiv:2509. 06296v2 Announce Type: replace-cross Abstract: Traditional on-policy reinforcement learning (RL) controllers for quadrupedal locomotion often suffer from low data efficiency, requiring millions of interactions with simulated environments to achieve stable control.
By Francisco Affonso, Felipe Tommaselli, Jo\~ao H. Al\'essio, Vivian S. Medeiros, Mateus V. Gasparino, Girish Chowdhary, Marcelo Becker
arXiv:2610.10297v1 Announce Type: cross
Abstract: While energy efficiency is a critical objective for legged-robot locomotion control, achieving low energy consumption while maintaining robust perfor...
By Ammar Issa, Anubhav Singh, Anton Tsaritsin, Sergey Kolyubin