A text-trajectory contrastive model with per-step cost assignment reduces safety violations in reinforcement learning agents under natural language constraints.
Safe reinforcement learning with natural language constraints,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
extension 1
citation-polarity summary
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1roles
extension 1polarities
extend 1representative citing papers
citing papers explorer
-
From Text to Trajectory: Exploring Complex Constraint Representation and Decomposition in Safe Reinforcement Learning
A text-trajectory contrastive model with per-step cost assignment reduces safety violations in reinforcement learning agents under natural language constraints.