How Well Do Latent World Models Understand Partially Observable Safety Constraints?

arXiv:2510.06492 · cs.RO · Submitted 2025-10-07 · Read on arXiv

cs.RO

Submitted: 2025-10-07

Updated: 2026-10-12

Project page: https://cmu-intentlab.github.io/multisafe

Terminology

• World Model — 5× in this paper · explained in 7 episodes
• Latent World Models — 2× in this paper · explained in 2 episodes
• Calibration — 1× in this paper · explained in 7 episodes
• Failure Modes — 1× in this paper · explained in 3 episodes
• Latent Representation — 1× in this paper · explained in 2 episodes

Sources

• World Models
• Evaluating Gemini Robotics Policies in a Veo World Simulator
• Guaranteed Discovery of Control-Endogenous Latent States with Multi-Step Inverse Models
• DINOv3

Related papers

• FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
• MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
• RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
• ARC: A Reasoning Recipe for Robot Foundation Models
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies