Pith. sign in

REVIEW 8 cited by

Learning to be Safe: Deep RL with a Safety Critic

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2010.14603 v1 pith:CB64UDTS submitted 2020-10-27 cs.LG cs.RO

classification cs.LGcs.RO
keywords learningsafetyenvironmentssafesystemstasksdeeplearn
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Safety is an essential component for deploying reinforcement learning (RL) algorithms in real-world scenarios, and is critical during the learning process itself. A natural first approach toward safe RL is to manually specify constraints on the policy's behavior. However, just as learning has enabled progress in large-scale development of AI systems, learning safety specifications may also be necessary to ensure safety in messy open-world environments where manual safety specifications cannot scale. Akin to how humans learn incrementally starting in child-safe environments, we propose to learn how to be safe in one set of tasks and environments, and then use that learned intuition to constrain future behaviors when learning new, modified tasks. We empirically study this form of safety-constrained transfer learning in three challenging domains: simulated navigation, quadruped locomotion, and dexterous in-hand manipulation. In comparison to standard deep RL techniques and prior approaches to safe RL, we find that our method enables the learning of new tasks and in new environments with both substantially fewer safety incidents, such as falling or dropping an object, and faster, more stable learning. This suggests a path forward not only for safer RL systems, but also for more effective RL systems.

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 8 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. SafeExplorer: An Unbiased Policy Gradient for Reinforcement Learning with Recovery Interventions

    cs.LG 2026-07 conditional novelty 6.0 of 10

    An unbiased PPO gradient that ignores recovery-policy density, plus analytic recovery values and success-gated imitation, cuts training falls by 26–233× on locomotion tasks without sacrificing reward.

  2. ARMOR: Robust Reinforcement Learning-based Control for UAVs under Physical Attacks

    cs.LG 2025-06 conditional novelty 6.0 of 10

    A teacher-student latent representation method lets an RL drone controller stay stable under GPS, gyroscope, and other sensor attacks, including attacks never seen in training.

  3. SafeMimic: Towards Safe and Autonomous Human-to-Robot Imitation for Mobile Manipulation

    cs.RO 2025-06 conditional novelty 6.0 of 10

    SafeMimic enables a mobile robot to safely and autonomously adapt a single third-person human video into a successful multi-step manipulation strategy.

  4. Approximating Safety Feedback Without a Safety Oracle via Model Predictive Control

    cs.LG 2025-10 conditional novelty 5.0 of 10

    RL-SA VMPC shields an RL policy by planning, via MPPI in a black-box simulator, a path from the next state back to the previous state, aborting when no such path exists.

  5. Verifiable Safety Q-Filters via Hamilton-Jacobi Reachability and Multiplicative Q-Networks

    cs.LG 2025-05 reject novelty 5.0 of 10

    Learned Q-function safety filters are certified by verifying two sufficient conditions with a mixed-integer optimizer, using a multiplicative Q-network to prevent safe-set collapse during fine-tuning.

  6. Learning Fast, Tool aware Collision Avoidance for Collaborative Robots

    cs.RO 2025-08 conditional novelty 4.0 of 10

    A real-time, tool-aware collision avoidance system for cobots that blends a learned perception-safety critic with classical IK, achieving low collision rates in dynamic partially-observed environments.

  7. Safe and Performant Controller Synthesis using Gradient-based Model Predictive Control and Control Barrier Functions

    eess.SY 2025-07 reject novelty 4.0 of 10

    A two-stage controller that uses L-BFGS gradient-based MPC for performance and a CBF-QP filter for hard safety constraints is demonstrated on simulated unicycle and planar quadrotor navigation.

  8. Safe and Performant Deployment of Autonomous Systems via Model Predictive Control and Hamilton-Jacobi Reachability Analysis

    cs.RO 2025-06 conditional novelty 3.0 of 10

    Adding a Hamilton-Jacobi reachability safety value as a terminal constraint in model predictive control makes the controller recursively feasible and reduces safety violations in car and robot arm simulations.

Pith tools