# XEWorld: Can Action-Conditioned World Models Generalize to Unseen Robot Embodiments?

arXiv: https://arxiv.org/abs/2608.05799 · HTML: https://mlsanity.com/paper/2608.05799/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Placed in Embodied and robot world models, one of 11 directions in World models.
- Builds on 18 of its 32 references that are in the map.
- First posted on arXiv 6 Aug 2026.

## Placed in

- World models / [Embodied and robot world models](https://mlsanity.com/wm/embodied-and-robot-world-models/)

## Builds on (most cited first, 18 of 18)

- [Wan: Open and Advanced Large-Scale Video Generative Models](https://mlsanity.com/paper/2503.20314/) (2025)
- [Cosmos World Foundation Model Platform for Physical AI](https://mlsanity.com/paper/2501.03575/) (2025)
- [RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation](https://mlsanity.com/paper/2506.18088/) (2025)
- [Towards World Simulator: Crafting Physical Commonsense-Based Benchmark for Video Generation](https://mlsanity.com/paper/2410.05363/) (2024)
- [Cosmos 3: Omnimodal World Models for Physical AI](https://mlsanity.com/paper/2606.02800/) (2026)
- [CoTracker3: Simpler and Better Point Tracking by Pseudo-Labelling Real Videos](https://mlsanity.com/paper/2410.11831/) (2024)
- [IRASim: A Fine-Grained World Model for Robot Manipulation](https://mlsanity.com/paper/2406.14540/) (2024)
- [WorldArena: A Unified Benchmark for Evaluating Perception and Functional Utility of Embodied World Models](https://mlsanity.com/paper/2602.08971/) (2026)
- [Being-H0.5: Scaling Human-Centric Robot Learning for Cross-Embodiment Generalization](https://mlsanity.com/paper/2601.12993/) (2026)
- [EnerVerse-AC: Envisioning Embodied Environments with Action Condition](https://mlsanity.com/paper/2505.09723/) (2025)
- [BridgeV2W: Bridging Video Generation Models to Embodied World Models via Embodiment Masks](https://mlsanity.com/paper/2602.03793/) (2026)
- [FlowWAM: Optical Flow as a Unified Action Representation for World Action Models](https://mlsanity.com/paper/2607.13017/) (2026)
- [Qwen-RobotWorld Technical Report: Unifying Embodied World Modeling through Language-Conditioned Video Generation](https://mlsanity.com/paper/2606.17030/) (2026)
- [OXE-AugE: A Large-Scale Robot Augmentation of OXE for Scaling Cross-Embodiment Policy Learning](https://mlsanity.com/paper/2512.13100/) (2025)
- [WorldArena 2.0: Extending Embodied World Model Benchmarking on Modality, Functionality and Platform](https://mlsanity.com/paper/2605.17912/) (2026)
- [Masked Visual Actions for Unified World Modeling](https://mlsanity.com/paper/2607.19343/) (2026)
- [AnyBody: A Benchmark Suite for Cross-Embodiment Manipulation](https://mlsanity.com/paper/2505.14986/) (2025)
- [EgoDemoGen: Egocentric Demonstration Generation for Viewpoint Generalization in Robotic Manipulation](https://mlsanity.com/paper/2509.22578/) (2025)

## Abstract (arXiv)

Action-conditioned world models are promising learned simulators for robotic manipulation, yet evaluating them exclusively on training robots fails to reveal whether they capture physical dynamics or merely memorize visual patterns. To answer whether a model can faithfully render a robot it has never seen, we introduce XEWorld, a controlled cross-embodiment testbed for world models that isolates embodiments by evaluating held-out robots within physically identical scenes. Our systematic analysis uncovers a shared architectural bottleneck: current models act primarily as 2D visual pattern matchers whose generalization is governed by visual similarity rather than physical kinematic similarity. Driven by this limitation, they struggle to translate abstract numeric joint actions into coherent visual trajectories, and fail to predict dynamic visual changes from static initial observations. Consequently, successfully rendering an unseen embodiment zero-shot strictly requires heavily grounded cues, specifically pixel-space actions and explicit spatial-temporal alignment. Even when bypassing this zero-shot barrier via few-shot adaptation, the forced appearance recovery triggers catastrophic forgetting of seen embodiments. Together, these failures expose a critical inability to apply learned physical dynamics to novel visual appearances, highlighting that achieving true cross-embodiment generalization requires architectural innovations that decouple visual appearance from underlying physical dynamics.