arXiv:2506. 01052v3 Announce Type: replace Abstract: We investigate the finite-time convergence properties of Temporal Difference (TD) learning with linear function approximation, a cornerstone of reinforcement learning.
arXiv:2606. 05967v1 Announce Type: cross Abstract: In this paper, we study the finite-time behavior of the TD(0) temporal-difference method with linear function approximation (LFA).
The paper introduces Fast Regularized Policy Mirror Descent (PMD) that pairs policy updates with a single temporal-difference (TD) critic step. It proves global linear convergence for finite discounted MDPs using exact coordinate-wise Bellman updates and any positive actor stepsize, regardless of critic initialization. For stochastic TD-PMD with strongly convex mirror maps, the authors achieve an expected value gap of ε after “~O(1/((1-γ)^5 σ_b ε))” transitions, without requiring trajectory resets, generative models, or nested policy evaluation loops.
In this paper, we study the finite-time behavior of the TD(0) temporal-difference method with linear function approximation (LFA). We consider on-policy independent and identically distributed (i.
arXiv:2609. 29961v1 Announce Type: new Abstract: Many iterative algorithms rely on bootstrapping.
The paper introduces a contraction framework for stochastic operators that incorporates bootstrapping, where a variable is updated using a frozen copy as a target that is refreshed every $K$ steps. By modeling the sampled update as a stochastic operator, the authors derive a finite‑time bound for i.i.d. samples that applies to any target‑update period and does not require gradient structure or uniformly bounded sampling error. The framework shows that the iterates converge geometrically in root mean square to a ball around the fixed point, with the error floor scaling with the step size, and it generalizes existing deterministic and stochastic‑gradient bounds.
arXiv:2606. 11431v1 Announce Type: new Abstract: Mirror Descent (MD) extends Gradient Descent (GD) beyond Euclidean geometry and has recently reappeared as a lens for KL-regularized policy optimization in reinforcement learning and LLM post-training.
arXiv:2605. 08488v2 Announce Type: replace-cross Abstract: We develop a unified Lyapunov-integral quadratic constraint (IQC) framework for establishing uniform stability of first-order accelerated optimization algorithms in the $\beta$-smooth and $\gamma$-strongly convex regime.
The paper introduces a finite‑dimensional sensitivity framework called the Learning Stability Profile, which captures how perturbations in inputs, parameter initialization, and update mechanisms propagate along a learning trajectory. It presents a Lyapunov criterion that, under regularity, coercivity, and dissipation assumptions, yields uniform or exponentially decaying bounds on linearized transition operators, distinguishing between terminal decay, boundedness, and subexponential growth. The framework is applied to various learning mechanisms—including feedforward networks, residual architectures, stochastic gradient methods, and locally Lipschitz systems—providing spectral, dissipativity, and contraction‑based stability bounds.
arXiv:2607. 22982v1 Announce Type: new Abstract: Natural Policy Gradient (NPG) is a well-established Reinforcement Learning algorithm that underlies widely used methods such as Trust Region Policy Optimization and Proximal Policy Optimization, both of which have demonstrated strong empirical success.
arXiv:2610. 14567v2 Announce Type: replace-cross Abstract: This paper develops strong Gaussian approximations (KMT type) for constant learning rate stochastic gradient descent (SGD).
arXiv:2608.22636v1 Announce Type: cross Abstract: Q-learning with linear function approximation can be unstable because an arbitrary approximation architecture need not preserve the Bellman contracti...