Choosing exploration policies by the ensemble disagreement of forward-backward value estimates improves zero-shot RL sample efficiency on DeepMind Control Suite tasks.
Radford M Neal
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
-
Epistemically-guided forward-backward exploration
Choosing exploration policies by the ensemble disagreement of forward-backward value estimates improves zero-shot RL sample efficiency on DeepMind Control Suite tasks.