mlsanityData to 7 Oct 2026

Do Androids Dream of Electric Fences? Safety-Aware Reinforcement Learning with Latent Shielding

arXiv 2112.11490 ↗First posted 21 Dec 2021cs.LG cs.LO cs.NEcited by 9

Not on the map. Why?

Abstract (arXiv)

The growing trend of fledgling reinforcement learning systems making their way into real-world applications has been accompanied by growing concerns for their safety and robustness. In recent years, a variety of approaches have been put forward to address the challenges of safety-aware reinforcement learning; however, these methods often either require a handcrafted model of the environment to be provided beforehand, or that the environment is relatively simple and low-dimensional. We present a novel approach to safety-aware deep reinforcement learning in high-dimensional environments called latent shielding. Latent shielding leverages internal representations of the environment learnt by model-based agents to "imagine" future trajectories and avoid those deemed unsafe. We experimentally demonstrate that this approach leads to improved adherence to formally-defined safety specifications.

Key facts. Cited by 9 papers in the map (data through 7 Oct 2026). Builds on 22 of its 25 references that are in the map. First posted on arXiv 21 Dec 2021.

Builds on

22 of 25 references.

References · 22

  1. 2019Dream to Control: Learning Behaviors by Latent Imaginationcited by 1,598
  2. 2019Mastering Atari, Go, Chess and Shogi by Planning with a Learned Modelcited by 1,327
  3. 2018Learning Latent Dynamics for Planning from Pixelscited by 1,221
  4. 2020Mastering Atari with Discrete World Modelscited by 990
  5. 2017Constrained Policy Optimizationcited by 928
  6. 2018Recurrent World Models Facilitate Policy Evolutioncited by 761
  7. 2017Safe Reinforcement Learning via Shieldingcited by 427
  8. 2015Risk-Constrained Reinforcement Learning with Percentile Risk Criteriacited by 320
  9. 2015Embed to Control: A Locally Linear Latent Dynamics Model for Control from Raw Imagescited by 266
  10. 2020Projection-Based Constrained Policy Optimizationcited by 195
  11. 2020Recovery RL: Safe Reinforcement Learning with Learned Recovery Zonescited by 156
  12. 2017Imagination-Augmented Agents for Deep Reinforcement Learningcited by 137
10 more
  1. 2020Learning to be Safe: Deep RL with a Safety Criticcited by 109
  2. 2020Conservative Safety Critics for Explorationcited by 107
  3. 2020Safe Reinforcement Learning via Curriculum Inductioncited by 73
  4. 2021Safe Multi-Agent Reinforcement Learning via Shieldingcited by 60
  5. 2020Neurosymbolic Reinforcement Learning with Formally Verified Explorationcited by 49
  6. 2019Robust Model Predictive Shielding for Safe Reinforcement Learning with Stochastic Dynamicscited by 45
  7. 2015From Pixels to Torques: Policy Learning with Deep Dynamical Modelscited by 44
  8. 2020Adaptive Shielding under Uncertaintycited by 17
  9. 2021Shielding Atari Games with Bounded Presciencecited by 14
  10. 2021Symbolic Reinforcement Learning for Safe RAN Controlcited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 1; 2023: 3; 2024: 1; 2025: 2; 2026: 2.

9 papers.

  1. 2023A Survey of Demonstration Learningcited by 12
  2. 2022Learning a Shield from Catastrophic Action Effects: Never Repeat the Same Mistakecited by 5
  3. 2024Leveraging Approximate Model-based Shielding for Probabilistic Safety Guarantees in Continuous Environmentscited by 5
  4. 2023Approximate Model-Based Shielding for Safe Reinforcement Learningcited by 4
  5. 2023Approximate Shielding of Atari Agents for Safe Explorationcited by 2
  6. 2026Robust Probabilistic Shielding for Safe Offline Reinforcement Learningcited by 2
  7. 2025Predictive Safety Shield for Dyna-Q Reinforcement Learningcited by 1
  8. 2026Safe Reinforcement Learning via Recovery-based Shielding with Gaussian Process Dynamics Modelscited by 1
  9. 2025Adaptive GR(1) Specification Repair for Liveness-Preserving Shielding in Reinforcement Learningcited by 0