mlsanityData to 14 Oct 2026

CRAX: Fast Safe Reinforcement Learning Benchmarking

arXiv 2606.20376 ↗First posted 18 Jun 2026cs.LG cs.AIcited by 2

On the map: Deep reinforcement learning methods → Safe and constrained RL

Abstract (arXiv)

Safety is a core concern for deploying reinforcement learning (RL) agents in real-world domains such as robotics and autonomous driving. While benchmarks have been central to progress in RL, existing 3D physics-based safety benchmarks remain computationally slow, limiting large-scale experimentation and rapid prototyping. To address this gap, we propose CRAX (Constrained RL Accelerated with JAX). Built on top of the MuJoCo XLA (MJX) physics engine, CRAX leverages vectorized operations and hardware acceleration, yielding up to 200x faster training over comparable CPU-based safety benchmarks. The benchmark features eight tasks spanning three difficulty levels and multiple agent morphologies. Evaluating seven popular safe RL methods, we find that none dominates across tasks, and that learning safe policies from pixels remains largely unsolved.

Key facts. Cited by 2 papers in the map (data through 14 Oct 2026). Placed in Safe and constrained RL, one of 15 directions in Deep reinforcement learning methods. Builds on 31 of its 53 references that are in the map. First posted on arXiv 18 Jun 2026.

Builds on

31 of 53 references.

Closest ancestors · 2

Same direction, newest first

  1. 2025SafeOR-Gym: A Benchmark Suite for Safe Reinforcement Learning Algorithms on Practical Operations Research Problems2 Jun
  2. 2024A Survey of Constraint Formulations in Safe Reinforcement Learning3 Feb

Other references · 29

  1. 2017Proximal Policy Optimization Algorithmscited by 15,215
  2. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,885
  3. 2017CARLA: An Open Urban Driving Simulatorcited by 2,449
  4. 2019Meta-World: A Benchmark and Evaluation for Multi-Task and Meta Reinforcement Learningcited by 1,204
  5. 2018DeepMind Control Suitecited by 956
  6. 2021Deep Reinforcement Learning at the Edge of the Statistical Precipicecited by 807
  7. 2020Image Augmentation Is All You Need: Regularizing Deep Reinforcement Learning from Pixelscited by 585
  8. 2021Brax -- A Differentiable Physics Engine for Large Scale Rigid Body Simulationcited by 407
  9. 2020Curriculum Learning for Reinforcement Learning Domains: A Framework and Surveycited by 373
  10. 2023Minigrid & Miniworld: Modular & Customizable Reinforcement Learning Environments for Goal-Oriented Taskscited by 327
  11. 2020Responsive Safety in Reinforcement Learning by PID Lagrangian Methodscited by 316
  12. 2021MetaDrive: Composing Diverse Driving Scenarios for Generalizable Reinforcement Learningcited by 290
17 more
  1. 2023Safety-Gymnasium: A Unified Safe Reinforcement Learning Benchmarkcited by 143
  2. 2020CRPO: A New Approach for Safe Reinforcement Learning with Convergence Guaranteecited by 134
  3. 2020Learning to Walk in the Real World with Minimal Human Effortcited by 111
  4. 2017AI Safety Gridworldscited by 110
  5. 2024Craftax: A Lightning-Fast Benchmark for Open-Ended Reinforcement Learningcited by 88
  6. 2025MuJoCo Playgroundcited by 78
  7. 2022Saute RL: Almost Surely Safe Reinforcement Learning Using State Augmentationcited by 69
  8. 2022Penalized Proximal Policy Optimization for Safe Reinforcement Learningcited by 68
  9. 2022VMAS: A Vectorized Multi-Agent Simulator for Collective Robot Learningcited by 68
  10. 2023JaxMARL: Multi-Agent RL Environments and Algorithms in JAXcited by 63
  11. 2023Jumanji: a Diverse Suite of Scalable Reinforcement Learning Environments in JAXcited by 58
  12. 2023XLand-MiniGrid: Scalable Meta-Reinforcement Learning Environments in JAXcited by 55
  13. 2021safe-control-gym: a Unified Benchmark Suite for Safe Learning-based Control and Reinforcement Learning in Roboticscited by 50
  14. 2021Direct Behavior Specification via Constrained Reinforcement Learningcited by 31
  15. 2021Measuring Sample Efficiency and Generalization in Reinforcement Learning Benchmarks: NeurIPS 2020 Procgen Benchmarkcited by 23
  16. 2025A Clean Slate for Offline Reinforcement Learningcited by 11
  17. 2025HASARD: A Benchmark for Vision-Based Safe Reinforcement Learning in Embodied Agentscited by 6

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2026: 2.

2 papers.

  1. 2026Evaluation Metrics for Safe Reinforcement Learningcited by 0
  2. 2026Factoriax: A GPU-Accelerated Factorio-Style Simulator for Reinforcement Learningcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Datasets and Benchmarks for Offline Safe Reinforcement Learning
  2. 2025Assuring the Safety of Reinforcement Learning Components: AMLAS-RL
  3. 2024A Dynamic Safety Shield for Safe and Efficient Reinforcement Learning of Navigation Tasks
  4. 2022A Review of Safe Reinforcement Learning: Methods, Theory and Applications
  5. 2024Safety through Permissibility: Shield Construction for Fast and Safe Reinforcement Learning
  6. 2025Safety Modulation: Enhancing Safety in Reinforcement Learning through Cost-Modulated Rewards
  7. 2023Safe Reinforcement Learning in a Simulated Robotic Arm
  8. 2025A Survey of Safe Reinforcement Learning and Constrained MDPs: A Technical Survey on Single-Agent and Multi-Agent Safety