1X releases the 1X World Model for humanoid policy evaluation and scaling

Evidence notes

• 1X shared progress on 1XWM, a world model that predicts the consequences of NEO's actions and accelerates policy evaluation without relying only on real-world physical testing.
• The post highlights action controllability, scaling results across tasks, and correlation between world-model forecasts and real-world robot outcomes.
• The release adds a first-party evaluation and training-system release to 1X's research history, not just a task demo.