EVAL learns the optimal policy for entropy-regularized average-reward MDPs by training neural networks to approximate the dominant eigenvector of a tilted transition matrix, with a variant that recovers the unregularized solution.
Title resolution pending
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
-
EVAL: EigenVector-based Average-reward Learning
EVAL learns the optimal policy for entropy-regularized average-reward MDPs by training neural networks to approximate the dominant eigenvector of a tilted transition matrix, with a variant that recovers the unregularized solution.