A reinforcement learning regularizer that maximizes trajectory-level total correlation produces simpler, more compressible policies that are more robust to perturbations.
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
-
Maximum Total Correlation Reinforcement Learning
A reinforcement learning regularizer that maximizes trajectory-level total correlation produces simpler, more compressible policies that are more robust to perturbations.