# X-Safe: Embodiment-Agnostic Safeguarding for Robot Arm Manipulation Policies via Action Masking

arXiv: https://arxiv.org/abs/2606.22278 · HTML: https://mlsanity.com/paper/2606.22278/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 34 of its 46 references that are in the map.
- First posted on arXiv 21 Jun 2026.

## Builds on (most cited first, 30 of 34)

- [π0: A Vision-Language-Action Flow Model for General Robot Control](https://mlsanity.com/paper/2410.24164/) (2024)
- [Safe Learning in Robotics: From Learning-Based Control to Safe Reinforcement Learning](https://mlsanity.com/paper/2108.06266/) (2021)
- [SafeVLA: Towards Safety Alignment of Vision-Language-Action Model via Constrained Learning](https://mlsanity.com/paper/2503.03480/) (2025)
- [Safe Reinforcement Learning via Projection on a Safe Set: How to Achieve Optimality?](https://mlsanity.com/paper/2004.00915/) (2020)
- [Verifiably Safe Exploration for End-to-End Reinforcement Learning](https://mlsanity.com/paper/2007.01223/) (2020)
- [Provably Safe Reinforcement Learning: Conceptual Analysis, Survey, and Benchmarking](https://mlsanity.com/paper/2205.06750/) (2022)
- [Reachability-based Trajectory Safeguard (RTS): A Safe and Fast Reinforcement Learning Safety Layer for Continuous Control](https://mlsanity.com/paper/2011.08421/) (2020)
- [Motion Planning around Obstacles with Convex Optimization](https://mlsanity.com/paper/2205.04422/) (2022)
- [Provably Safe Reinforcement Learning via Action Projection using Reachability Analysis and Polynomial Zonotopes](https://mlsanity.com/paper/2210.10691/) (2022)
- [Sampling-Based Motion Planning: A Comparative Review](https://mlsanity.com/paper/2309.13119/) (2023)
- [Value Alignment Verification](https://mlsanity.com/paper/2012.01557/) (2020)
- [Excluding the Irrelevant: Focusing Reinforcement Learning through Continuous Action Masking](https://mlsanity.com/paper/2406.03704/) (2024)
- [Provably Safe Deep Reinforcement Learning for Robotic Manipulation in Human Environments](https://mlsanity.com/paper/2205.06311/) (2022)
- [Safety Filtering While Training: Improving the Performance and Sample Efficiency of Reinforcement Learning Agents](https://mlsanity.com/paper/2410.11671/) (2024)
- [DDAT: Diffusion Policies Enforcing Dynamically Admissible Robot Trajectories](https://mlsanity.com/paper/2502.15043/) (2025)
- [Safe, Task-Consistent Manipulation with Operational Space Control Barrier Functions](https://mlsanity.com/paper/2503.06736/) (2025)
- [APEX: Ambidextrous Dual-Arm Robotic Manipulation Using Collision-Free Generative Diffusion Models](https://mlsanity.com/paper/2404.02284/) (2024)
- [Dynamic Rank Adjustment in Diffusion Policies for Efficient and Flexible Training](https://mlsanity.com/paper/2502.03822/) (2025)
- [MIMIC-D: Multi-modal Imitation for MultI-agent Coordination with Decentralized Diffusion Policies](https://mlsanity.com/paper/2509.14159/) (2025)
- [Safe Reinforcement Learning using Data-Driven Predictive Control](https://mlsanity.com/paper/2211.11027/) (2022)
- [POLICEd RL: Learning Closed-Loop Robot Control Policies with Provable Satisfaction of Hard Constraints](https://mlsanity.com/paper/2403.13297/) (2024)
- [Efficient Motion Planning for Manipulators with Control Barrier Function-Induced Neural Controller](https://mlsanity.com/paper/2404.01184/) (2024)
- [RAIL: Reachability-Aided Imitation Learning for Safe Policy Execution](https://mlsanity.com/paper/2409.19190/) (2024)
- [Safe Reinforcement Learning using Action Projection: Safeguard the Policy or the Environment?](https://mlsanity.com/paper/2509.12833/) (2025)
- [V-OCBF: Learning Safety Filters from Offline Data via Value-Guided Offline Control Barrier Functions](https://mlsanity.com/paper/2512.10822/) (2025)
- [BiCoord: A Bimanual Manipulation Benchmark towards Long-Horizon Spatial-Temporal Coordination](https://mlsanity.com/paper/2604.05831/) (2026)
- [Faster Algorithms for Growing Collision-Free Convex Polytopes in Robot Configuration Space](https://mlsanity.com/paper/2410.12649/) (2024)
- [A General Safety Framework for Autonomous Manipulation in Human Environments](https://mlsanity.com/paper/2412.10180/) (2024)
- [Neural Configuration-Space Barriers for Manipulation Planning and Control](https://mlsanity.com/paper/2503.04929/) (2025)
- [Superfast Configuration-Space Convex Set Computation on GPUs for Online Motion Planning](https://mlsanity.com/paper/2504.10783/) (2025)

## Built on by (most cited first, 2 of 2)

- [CrossSafe: Towards Cross-Embodiment Latent Safety Filters](https://mlsanity.com/paper/2609.28984/) (2026)
- [Multi-Link Safety Filtering for VLA Policies Around Moving Hazards](https://mlsanity.com/paper/2609.40007/) (2026)

## Abstract (arXiv)

Ensuring safety of learning-enabled robotic manipulation across diverse embodiments and tasks still requires significant manual engineering. For collision avoidance, existing approaches typically rely on embodiment-specific learned safety approximations or simplified robot models together with stopping-based fallback controllers that compromise performance. In this paper, we propose an embodiment-agnostic safeguarding approach, X-Safe, which provides formal probabilistic guarantees for collision avoidance without model simplifications or learned safety functions. By operating in the configuration space, our safeguarding method works with different types of robot policies. X-Safe reasons locally around the current configuration to establish safe regions and uses action masking to restrict the policy's support to these safe regions. We demonstrate X-Safe on two embodiments and three policies, both in simulation and on hardware. We observe less than half of the performance degradation compared to state-of-the-art safeguarding, and empirically corroborate our formal collision avoidance guarantees.