arXiv research

This work explores representation complexity in RL paradigms, revealing model-based RL as the easiest task.

This paper surveys RL for combinatorial optimization, focusing on TSP.

Optimistic RL algorithms are simplified for deep RL with competitive performance.

Automates RL with sample-efficient hyperparameter optimization.

Paper establishes lower bounds and optimal algorithms for deployment-efficient RL.

RH-UCRL combines pessimism and optimism for robust RL.

M3PO improves model-based meta-RL with theoretical guarantees.

MOReL learns offline RL policies using pessimistic MDPs.

This study explains RL training dynamics in LLMs, focusing on token-level optimization and reasoning pattern reshaping.

Myopic optimization outperforms reinforcement learning in portfolio management, leading to lower returns and higher risks.

New RL algorithm minimizes distributional learning error.

Optimistic NPG improves policy optimization in online RL with efficient sample complexity.

Offline RL policies should adapt to unknown aspects of the environment.

Paper connects RL and non-equilibrium statistical mechanics for entropy-regularized RL.

Unified framework for CO problems using RL, providing optimal solutions and convergence guarantees.

ContraBAR uses contrastive learning to learn Bayes-optimal policies in RL.

Unified hybrid RL algorithm improves online RL performance with offline data.

RL framework optimizes trading costs in noisy markets.

Experimentally, it has been observed that humans and animals often make decisions that do not maximize their expected utility, but rather choose outcomes randomly, with probability proportional to expected utility. Probability matching, as this strategy is called, is equivalent to maximum entropy reinforcement learning…

Reinforcement learning (RL) methods often rely on massive exploration data to search optimal policies, and suffer from poor sampling efficiency. This paper presents a mixed reinforcement learning (mixed RL) algorithm by simultaneously using dual representations of environmental dynamics to search the optimal policy wit…

Critic-regularized regression improves offline RL performance.

Survey of RL methods for optimizing power grid topologies.

New offline RL algorithms tackle partial data coverage with optimal performance and practicality.

Develops ODRPO to improve RL algorithms with better performance and stability.

To overcome the curses of dimensionality and modeling of Dynamic Programming (DP) methods to solve Markov Decision Process (MDP) problems, Reinforcement Learning (RL) methods are adopted in practice. Contrary to traditional RL algorithms which do not consider the structural properties of the optimal policy, we propose …

New algorithm reduces offline RL data requirements significantly.

To overcome the curse of dimensionality and curse of modeling in Dynamic Programming (DP) methods for solving classical Markov Decision Process (MDP) problems, Reinforcement Learning (RL) algorithms are popular. In this paper, we consider an infinite-horizon average reward MDP problem and prove the optimality of the th…

A new offline RL framework unifies imitation learning and vanilla offline RL.

New algorithm for offline RL with linear approx in MDPs and MGs, nearly optimal.

This paper bridges offline and online RL by studying policy finetuning with a reference policy.

HyperAgent improves RL exploration in large-scale problems.

EQO uses a simple bonus term for efficient exploration in tabular RL.

Multi-task reinforcement learning (RL) aims to simultaneously learn policies for solving many tasks. Several prior works have found that relabeling past experience with different reward functions can improve sample efficiency. Relabeling methods typically ask: if, in hindsight, we assume that our experience was optimal…

RL improves combinatorial optimization by automating heuristic search.

Although reinforcement learning (RL) can provide reliable solutions in many settings, practitioners are often wary of the discrepancies between the RL solution and their status quo procedures. Therefore, they may be reluctant to adapt to the novel way of executing tasks proposed by RL. On the other hand, many real-worl…

Reinforcement learning (RL) algorithms have been successfully applied to a range of challenging sequential decision making and control tasks. In this paper, we classify RL into direct and indirect RL according to how they seek the optimal policy of the Markov decision process problem. The former solves the optimal poli…

Mirror descent method improved RL algorithms.

Fine-tuning RL with offline data reduces online interactions.

MOPO optimizes offline RL by penalizing dynamics uncertainty.

This tutorial reviews RL-based methods for optimizing diffusion models to maximize specific metrics.

ReSkill reconciles RL skill creation with policy optimization.

FMDP-BF algorithm improves RL in factored MDPs with exponential regret reduction.

Hybrid RL algorithms improve offline and online RL in linear MDPs.

Proposes RFQI for robust RL using offline data.

Designing optimal controllers continues to be challenging as systems are becoming complex and are inherently nonlinear. The principal advantage of reinforcement learning (RL) is its ability to learn from the interaction with the environment and provide optimal control strategy. In this paper, RL is explored in the cont…

Paper tackles RL for power grid topology optimization.

New RL algorithm tackles online robust MDPs with uncertainty.

Paper proposes an RL algorithm to ensure policy performance guarantees.