(40) Here, H(·) is the entropy function satisfying 0 ⩽ H(p) ⩽ log |A|, ∀p ∈ ∆(A)

It follows that πh ∈ Ph, for all s ∈ S, πh(·|s) is the solution to the following optimization problem [Beck · 2017

1 Pith paper cite this work. Polarity classification is still indexing.

1 Pith paper citing it

browse 1 citing papers

citation-role summary

method 1

use method 1

cs.LG · 2025-06-27 · conditional · novelty 6.0

VAC is a new actor-critic method with a single optimistic objective and a provably near-optimal regret bound in linear Markov decision processes.

Showing 1 of 1 citing paper.

Exploration from a Primal-Dual Lens: Value-Incentivized Actor-Critic Methods for Sample-Efficient Online RL cs.LG · 2025-06-27 · conditional · none · ref 3
VAC is a new actor-critic method with a single optimistic objective and a provably near-optimal regret bound in linear Markov decision processes.