Action aliasing from safety projections harms policy-gradient estimates more severely when the projection is inside the policy than when it is outside, but a penalty term restores competitiveness.
The environment has the state x= [ ϑ,˙ϑ ]T and the dynamics ˙x= ( ˙ϑ g ℓsin(ϑ) +1 mℓ2u ) ,(53) wheregis gravity andm,ℓare the mass and the length of the pendulum, respectively
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Safe Reinforcement Learning using Action Projection: Safeguard the Policy or the Environment?
Action aliasing from safety projections harms policy-gradient estimates more severely when the projection is inside the policy than when it is outside, but a penalty term restores competitiveness.