NBDI learns skill termination from state-action novelty (ICM prediction error) on task-agnostic demonstrations, improving downstream RL performance in maze and manipulation benchmarks.
In the maze and sparse block stacking environment, a single fully-connected layer with hidden dimension 52 and 70 have been used, 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
-
NBDI: A Simple and Effective Termination Condition for Skill Extraction from Task-Agnostic Demonstrations
NBDI learns skill termination from state-action novelty (ICM prediction error) on task-agnostic demonstrations, improving downstream RL performance in maze and manipulation benchmarks.