An adaptive reward-shaping method that periodically inflates distance-to-acceptance values for under-progressed stages lets RL agents reach the best achievable task progression on co-safe LTL tasks within a finite number of updates.
Ltl and beyond: Formal languages for reward function specification in reinforcement learning
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.RO 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Adaptive Reward Design for Reinforcement Learning
An adaptive reward-shaping method that periodically inflates distance-to-acceptance values for under-progressed stages lets RL agents reach the best achievable task progression on co-safe LTL tasks within a finite number of updates.