REVIEW 6 cited by
Constrained Policy Optimization via Bayesian World Models
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
Signed reviews
read the original abstract
Improving sample-efficiency and safety are crucial challenges when deploying reinforcement learning in high-stakes real world applications. We propose LAMBDA, a novel model-based approach for policy optimization in safety critical tasks modeled via constrained Markov decision processes. Our approach utilizes Bayesian world models, and harnesses the resulting uncertainty to maximize optimistic upper bounds on the task objective, as well as pessimistic upper bounds on the safety constraints. We demonstrate LAMBDA's state of the art performance on the Safety-Gym benchmark suite in terms of sample efficiency and constraint violation.
Forward citations
Cited by 6 Pith papers
-
Learning Safe Agent Behaviour from Human Preferences and Justifications via World Models
DROPJ trains a world-model-based MPC agent from one-shot human preferences plus safety justifications, cutting training cost and improving deployment safety in car-racing simulations.
-
Safe Planning and Policy Optimization via World Model Learning
SPOWL is a model-based safe RL method that uses a value-equivalent world model, a Lagrangian-trained safe policy, and adaptive planning thresholds to achieve low-cost, high-reward control on SafetyGymnasium tasks.
-
Safe Reinforcement Learning using Finite-Horizon Gradient-based Estimation
The paper introduces Gradient-based Estimation (GBE), a Taylor-expansion method using analytic trajectory gradients, and the trust-region algorithm CGPO for safe RL with finite-horizon constraints.
-
Q-learning-based Model-free Safety Filter
A Q-learning safety filter with a time-dependent reward blocks unsafe actions from arbitrary task policies, but its theoretical guarantee is not valid as written.
-
ACORN: Adaptive Contrastive Optimization for Safe and Robust Fine-Grained Robotic Manipulation
ACORN adds contrastive learning with action-space perturbations to ACT imitation learning and reports improved safety metrics on one bimanual task, but the headline metric is undefined and results are inconsistent acr...
-
Combining Bayesian Inference and Reinforcement Learning for Agent Decision Making: A Review
A survey that organizes combinations of Bayesian inference and reinforcement learning, rates them on four properties, and raises ten open questions.
Discussion (0). Continue with ORCID to comment.