PhysisForcing: Physics Reinforced World Simulator for Robotic Manipulation

Hugging Face Daily Papers Papers
robotic-manipulation video-generation physical-consistency world-simulator embodied-ai diffusion-model dit

Summary

PhysisForcing is a training framework that enhances embodied video generation for robotic manipulation by enforcing physical consistency through pixel-level trajectory alignment and semantic-level relational alignment losses in a DiT-based architecture, achieving notable improvements on benchmarks.

Video generation models have emerged as a promising paradigm for embodied world simulation. However, both general-domain video generators and robot-specific data fine-tuned models can still produce physically implausible manipulations, including discontinuous motion trajectories and inconsistent robot-object interactions, which limits their reliability as world simulators. Through extensive experiments, we find that such physical instability mainly arises from two factors: deformation of moving objects and implausible spatio-temporal correlations among interacting entities, particularly during contact. Building on this observation, we propose PhysisForcing, a scalable training framework that strengthens physical consistency by focusing supervision on physics-informative regions through joint optimization of pixel-level and semantic-level features. The framework consists of a pixel-level trajectory alignment loss, which supervises DiT features using reference point trajectories, and a semantic-level relational alignment loss, which aligns DiT features with inter-region relations extracted from a frozen video understanding encoder. Extensive experiments on R-Bench, PAI-Bench, and EZS-Bench show that PhysisForcing consistently improves embodied video generation over strong baselines, improving the Wan2.2-I2V-A14B and Cosmos3-Nano base models on R-Bench by 22.3\% and 9.2\% (7.1\% and 3.7\% over vanilla finetuning), with the Cosmos3-Nano variant attaining the best overall score. Beyond generation, as a world model under the WorldArena action-planner protocol it raises the closed-loop success rate from 16.0\% to 24.0\% and further improves downstream policy success, indicating that physically aligned video models yield stronger representations for robotic manipulation.
Original Article
View Cached Full Text

Cached at: 06/29/26, 02:00 AM

Paper page - PhysisForcing: Physics Reinforced World Simulator for Robotic Manipulation

Source: https://huggingface.co/papers/2606.28128 Published on Jun 26

#1 Paper of the day Authors:

,

,

,

,

,

,

,

,

,

,

,

,

Abstract

PhysisForcing enhances embodied video generation by enforcing physical consistency through pixel-level trajectory alignment and semantic-level relational alignment losses in a DiT-based framework.

Video generation modelshave emerged as a promising paradigm forembodied world simulation. However, both general-domain video generators and robot-specific data fine-tuned models can still produce physically implausible manipulations, including discontinuous motion trajectories and inconsistent robot-object interactions, which limits their reliability as world simulators. Through extensive experiments, we find that such physical instability mainly arises from two factors: deformation of moving objects and implausible spatio-temporal correlations among interacting entities, particularly during contact. Building on this observation, we propose PhysisForcing, a scalable training framework that strengthensphysical consistencyby focusing supervision on physics-informative regions through joint optimization of pixel-level and semantic-level features. The framework consists of apixel-level trajectory alignment loss, which supervisesDiT featuresusing reference point trajectories, and asemantic-level relational alignment loss, which alignsDiT featureswith inter-region relations extracted from a frozenvideo understanding encoder. Extensive experiments onR-Bench,PAI-Bench, andEZS-Benchshow that PhysisForcing consistently improves embodied video generation over strong baselines, improving the Wan2.2-I2V-A14B and Cosmos3-Nano base models onR-Benchby 22.3\% and 9.2\% (7.1\% and 3.7\% over vanilla finetuning), with the Cosmos3-Nano variant attaining the best overall score. Beyond generation, as a world model under theWorldArenaaction-planner protocolit raises theclosed-loop success ratefrom 16.0\% to 24.0\% and further improves downstream policy success, indicating that physically aligned video models yield stronger representations forrobotic manipulation.

View arXiv pageView PDFProject pageAdd to collection

Get this paper in your agent:

hf papers read 2606\.28128

Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash

Models citing this paper0

No model linking this paper

Cite arxiv.org/abs/2606.28128 in a model README.md to link it from this page.

Datasets citing this paper0

No dataset linking this paper

Cite arxiv.org/abs/2606.28128 in a dataset README.md to link it from this page.

Spaces citing this paper0

No Space linking this paper

Cite arxiv.org/abs/2606.28128 in a Space README.md to link it from this page.

Collections including this paper0

No Collection including this paper

Add this paper to acollectionto link it from this page.

Similar Articles

PhysiFormer: Learning to Simulate Mechanics in World Space

Hugging Face Daily Papers

PhysiFormer uses coordinate-space diffusion to generate physically-plausible 3D object motions without explicit inductive biases, enabling efficient multi-object reasoning and generalization to complex materials and geometries.