WALL-WM 2.0 RoboDojo Preview

Official version
Papers and technical reports

X Square Robot’s WALL-WM 2.0 world-action model preview is the pretrained starting point for RoboDojo training. It reuses the Qwen3.5-9B tokenizer and chat template and provides the model configuration and normalization statistics for actions and states. The later full release differs mainly in the number of training iterations.

X Square Robot released the pretrained weights for the RoboDojo preview. The model repository also provides the configuration and action and state normalization files.

Papers and technical reports

Paper / report unconfirmed

Official introductory materials are recorded; no version-specific paper or technical report has been verified.

Last checked: 2026-11-21

How to access and use

Model files
Public access

Training code for this version is unconfirmed. Public code resources do not establish access to the exact training implementation.

Training / fine-tuning code

Model files

Model files
RoboDojo preview pretrained weightsAccessWeights license
Release information

SourceVerified on · 2026-11-21

Checks before deployment

Available weights or code do not establish plug-and-play robot compatibility. Check the required configuration, data and interfaces in the source documentation.

Deployment and usage resources

Architecture and inputs / outputs

Model type
Action policy
Architecture
WAM
Backbone
Wan2.2-TI2V-5B
Input
Language / RGB images / Robot state
Released
November 21, 2026
  • Backbone · Input modalities · Main category · Architecture: X Square Robot
  • Input modalities · Main category · Architecture: X Square Robot

Scope and training

Relative Cartesian end-effector positions, 6D rotations, and gripper actions for both arms.

Functional purpose

action_control

Record identity

Official version

Training

Pretraining

  • Functional purpose · Record identity · Training: X Square Robot
  • Relative Cartesian end-effector positions, 6D rotations, and gripper actions for both arms. · continuous: X Square Robot
Submit a source or correction