We study policy optimization for infinite-horizon, discounted constrained Markov decision processes. While primal-dual (PD) methods are widely used in practice, their theoretical guarantees typically hold for an impractical mixture policy. In contrast, dual descent (DD) methods output a single policy ("last iterate") with guarantees on its performance.
However, such methods require iteratively solving entropy-regularized MDPs in an inner-loop, limiting their practical effectiveness. Moreover, existing analyses rely on strong coverage assumptions that are difficult to satisfy in practice.
Consequently, we first provide a new analysis of DD methods that does not require coverage assumptions. We then employ entropy-regularized natural policy gradient and dual regularization to eliminate the inner loop, resulting in an efficient PD algorithm.
We develop a novel tracking analysis which shows that, compared to existing methods, the proposed algorithm achieves improved last-iterate convergence guarantees. Furthermore, our theoretical framework allows us to analyze a truncated bilevel method that interpolates between DD and PD methods.
Finally, we generalize our analysis to handle linear function approximation. Compared to prior last-iterate methods, we prove stronger guarantees under less restrictive assumptions.