GPWM: Modeling the World as 3D Particle Motion
GPWM: Modeling the World as 3D Particle Motion
This research presents GPWM, a generative particle world model that predicts how 3D scenes evolve by generating the motion of particles and rigid bodies over time. Unlike video-based world models that focus on appearance, GPWM operates directly in 3D physical space using a unified architecture for rigid, deformable, cloth, and granular materials. The model enables contact-rich planning by learning a generative prior over joint actuator-object motion, allowing open-loop execution in simulation for goal-directed manipulation tasks like pushing and picking without reference action trajectories.
The paper introduces GPWM, a generative particle world model that predicts 3D scene evolution by modeling particle and rigid body motion using flow matching and diffusion forcing, evaluated on a 25.5-million-frame corpus across six physical domains with reported prediction error reductions.
Source: arXiv Robotics — research abstracts · Read original article ↗
Loading article text…
What the source reports
Publisher-reported claims, with original evidence. These results have not been independently verified by RoboSignal.
- Environment
- SimulationOpen source S6
- Data origin
- Robot trajectoriesOpen source S3
Reported numbers
training frames
25.5
View original evidence
ysics and robotic interaction. GPWM is an object-aware multi-resolution transformer that combines flow matching with diffusion forcing to generate coherent physical trajectories far beyond its training horizon. We train and evaluate GPWM on ParticleWorlds, a 25.5-million-frame corpus spanning diverse geometries, materials, interactions, and robot embodiments. GPWM outperforms the best prior learned-dynamics baseline
Open source S3physical domains
6
View original evidence
in each of six physical domains, reducing prediction error by 31% on average. Breaking from the domain-specific setting of prior methods, a single Unified GPWM models these heterogeneous regimes with one set of weights, largely outperforming per-domain baselines and even surpassing our own specialist models in two domains. GPWM further generalizes to unseen geometries, physical properties, object counts, and rollout
Open source S4
- paper: not_reported
- code: not_reported
- weights: not_reported
- dataset: not_reported
Source excerpts and review record
Automatically extracted; no manual editorial approval recorded.
sequences are hard to discover. Executed open-loop in simulation, the resulting plans achieve goal-directed manipulation without reference action trajectories or goal-specific policies.
Open source S6
ysics and robotic interaction. GPWM is an object-aware multi-resolution transformer that combines flow matching with diffusion forcing to generate coherent physical trajectories far beyond its training horizon. We train and evaluate GPWM on ParticleWorlds, a 25.5-million-frame corpus spanning diverse geometries, materials, interactions, and robot embodiments. GPWM outperforms the best prior learned-dynamics baseline
Open source S3
in each of six physical domains, reducing prediction error by 31% on average. Breaking from the domain-specific setting of prior methods, a single Unified GPWM models these heterogeneous regimes with one set of weights, largely outperforming per-domain baselines and even surpassing our own specialist models in two domains. GPWM further generalizes to unseen geometries, physical properties, object counts, and rollout
Open source S4
horizons. Beyond prediction, GPWM enables contact-rich planning by learning a generative prior over joint actuator-object motion. Instead of searching over unconstrained actions, planning searches the model's latent noise for interactions that achieve a desired object-space goal. With the same world model, noise-space planning substantially outperforms action-space search on pushing and picking, where complex contact
Open source S5
Implications for data suppliers
RoboSignal interpretation and collection questions, not statements of buyer demand.
- Match the buyer’s robot embodiment, action space, sensors and timestamp requirements before planning collection.
- Compare the reported units and scope before using these quantities in a budget. Recording hours, sensor-hours and trajectories are different measures.
- Validate demand and acceptance criteria with a buyer before scaling. Publication, popularity and a research result do not establish a purchase commitment.
Source：arXiv Robotics — research abstracts · arxiv.org