Under mu-resets and policy realizability, sample complexity is exp(Theta(H)) under all-policy concentrability and exp(Theta(sqrt H)) under pushforward concentrability.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
The Sample Complexity of Policy Learning with Mu-Resets
Under mu-resets and policy realizability, sample complexity is exp(Theta(H)) under all-policy concentrability and exp(Theta(sqrt H)) under pushforward concentrability.