Reinforcement learning with outcome-based feedback faces a fundamental challenge: when rewards are only observed at trajectory endpoints, how do we assign credit to the right actions? This paper provides the first comprehensive analysis of this problem in online RL with general function approximation. We develop a provably sample-efficient algorithm achieving $\widetilde{O}({C_{\rm cov} H^3}/{ε^2})$ sample complexity, where $C_{\rm cov}$ is the coverability coefficient of the underlying MDP. By leveraging general function approximation, our approach works effectively in large or infinite state spaces where tabular methods fail, requiring only that value functions and reward functions can be represented by appropriate function classes. Our results also characterize when outcome-based feedback is statistically separated from per-step rewards, revealing an unavoidable exponential separation for certain MDPs. For deterministic MDPs, we show how to eliminate the completeness assumption, dramatically simplifying the algorithm. We further extend our approach to preference-based feedback settings, proving that equivalent statistical efficiency can be achieved even under more limited information. Together, these results constitute a theoretical foundation for understanding the statistical properties of outcome-based reinforcement learning.

This paper investigates the application of Reinforcement Learning (RL) to optimise call routing in call centres to minimise client waiting time and staff idle time. Two methods are compared: a model-based approach using Value Iteration (VI) under known system dynamics, and a model-free approach using Proximal Policy Optimisation (PPO) that learns from experience. For the model-based approach, a theoretical model is used, while a simulation model combining Discrete Event Simulation (DES) with the OpenAI Gym environment is developed for model-free learning. Both models frame the problem as a Markov Decision Process (MDP) within a Skills-Based Routing (SBR) framework, with Poisson client arrivals and exponentially distributed service and abandonment times. For policy evaluation, random, VI, and PPO policies are evaluated using the simulation model. After 1,000 test episodes, PPO consistently achives the highest rewards, along with the lowest client waiting time and staff idle time, despite requiring longer training time.

The rise of process data availability has recently led to the development of data-driven learning approaches. However, most of these approaches restrict the use of the learned model to predict the future of ongoing process executions. The goal of this paper is moving a step forward and leveraging available data to learning to act, by supporting users with recommendations derived from an optimal strategy (measure of performance). We take the optimization perspective of one process actor and we recommend the best activities to execute next, in response to what happens in a complex external environment, where there is no control on exogenous factors. To this aim, we investigate an approach that learns, by means of Reinforcement Learning, the optimal policy from the observation of past executions and recommends the best activities to carry on for optimizing a Key Performance Indicator of interest. The validity of the approach is demonstrated on two scenarios taken from real-life data.

Cooperative multi-agent reinforcement learning (MARL) is typically formalised as a Decentralised Partially Observable Markov Decision Process (Dec-POMDP), where agents must reason about the environment and other agents' behaviour. In practice, current model-free MARL algorithms use simple recurrent function approximators to address the challenge of reasoning about others using partial information. In this position paper, we argue that the empirical success of these methods is not due to effective Markov signal recovery, but rather to learning simple conventions that bypass environment observations and memory. Through a targeted case study, we show that co-adapting agents can learn brittle conventions, which then fail when partnered with non-adaptive agents. Crucially, the same models can learn grounded policies when the task design necessitates it, revealing that the issue is not a fundamental limitation of the learning models but a failure of the benchmark design. Our analysis also suggests that modern MARL environments may not adequately test the core assumptions of Dec-POMDPs. We therefore advocate for new cooperative environments built upon two core principles: (1) behaviours grounded in observations and (2) memory-based reasoning about other agents, ensuring success requires genuine skill rather than fragile, co-adapted agreements.

Reinforcement learning (RL) has achieved remarkable success in fields like robotics and autonomous driving, but adversarial attacks designed to mislead RL systems remain challenging. Existing approaches often rely on modifying the environment or policy, limiting their practicality. This paper proposes an adversarial attack method in which existing agents in the environment guide the target policy to output suboptimal actions without altering the environment. We propose a reward iteration optimization framework that leverages large language models (LLMs) to generate adversarial rewards explicitly tailored to the vulnerabilities of the target agent, thereby enhancing the effectiveness of inducing the target agent toward suboptimal decision-making. Additionally, a critical state identification algorithm is designed to pinpoint the target agent's most vulnerable states, where suboptimal behavior from the victim leads to significant degradation in overall performance. Experimental results in diverse environments demonstrate the superiority of our method over existing approaches.

In this paper, we tackle the challenge of radio access network (RAN) slicing within an open RAN (O-RAN) architecture. Our focus centers on a network that includes multiple mobile virtual network operators (MVNOs) competing for physical resource blocks (PRBs) with the goal of meeting probabilistic delay upper bound constraints for their clients while minimizing PRB utilization. Initially, we derive a reward function based on the law of large numbers (LLN), then implement practical modifications to adapt it for real-world experimental scenarios. We then propose our solution, the Percentile-based Delay-Aware Deep Reinforcement Learning (PDA-DRL), which demonstrates its superiority over several baselines, including DRL models optimized for average delay constraints, by achieving a 38\% reduction in resultant average delay. Furthermore, we delve into the issue of model weight sharing among multiple MVNOs to develop a robust personalized model. We introduce a reward-based personalization method where each agent prioritizes other agents' model weights based on their performance. This technique surpasses traditional aggregation methods, such as federated averaging, and strategies reliant on traffic patterns and model weight distance similarities.

Video Temporal Grounding (VTG) aims to localize relevant temporal segments in videos given natural language queries. Despite recent progress with large vision-language models (LVLMs) and instruction-tuning, existing approaches often suffer from limited temporal awareness and poor generalization. In this work, we introduce a two-stage training framework that integrates supervised fine-tuning with reinforcement learning (RL) to improve both the accuracy and robustness of VTG models. Our approach first leverages high-quality curated cold start data for SFT initialization, followed by difficulty-controlled RL to further enhance temporal localization and reasoning abilities. Comprehensive experiments on multiple VTG benchmarks demonstrate that our method consistently outperforms existing models, particularly in challenging and open-domain scenarios. We conduct an in-depth analysis of training strategies and dataset curation, highlighting the importance of both high-quality cold start data and difficulty-controlled RL. To facilitate further research and industrial adoption, we release all intermediate datasets, models, and code to the community.

Due to practical constraints such as partial observability and limited communication, Centralized Training with Decentralized Execution (CTDE) has become the dominant paradigm in cooperative Multi-Agent Reinforcement Learning (MARL). However, existing CTDE methods often underutilize centralized training or lack theoretical guarantees. We propose Multi-Agent Guided Policy Optimization (MAGPO), a novel framework that better leverages centralized training by integrating centralized guidance with decentralized execution. MAGPO uses an auto-regressive joint policy for scalable, coordinated exploration and explicitly aligns it with decentralized policies to ensure deployability under partial observability. We provide theoretical guarantees of monotonic policy improvement and empirically evaluate MAGPO on 43 tasks across 6 diverse environments. Results show that MAGPO consistently outperforms strong CTDE baselines and matches or surpasses fully centralized approaches, offering a principled and practical solution for decentralized multi-agent learning. Our code and experimental data can be found in https://github.com/liyheng/MAGPO.

We introduce a reinforcement learning (RL) based adaptive optimization algorithm for aerodynamic shape optimization focused on dimensionality reduction. The form in which RL is applied here is that of a surrogate-based, actor-critic policy evaluation MCMC approach allowing for temporal 'freezing' of some of the parameters to be optimized. The goals are to minimize computational effort, and to use the observed optimization results for interpretation of the discovered extrema in terms of their role in achieving the desired flow-field. By a sequence of local optimized parameter changes around intermediate CFD simulations acting as ground truth, it is possible to speed up the global optimization if (a) the local neighbourhoods of the parameters in which the changed parameters must reside are sufficiently large to compete with the grid-sized steps and its large number of simulations, and (b) the estimates of the rewards and costs on these neighbourhoods necessary for a good step-wise parameter adaption are sufficiently accurate. We give an example of a simple fluid-dynamical problem on which the method allows interpretation in the sense of a feature importance scoring.

Contemporary database systems, while effective, suffer severe issues related to complexity and usability, especially among individuals who lack technical expertise but are unfamiliar with query languages like Structured Query Language (SQL). This paper presents a new database system supported by Artificial Intelligence (AI), which is intended to improve the management of data using natural language processing (NLP) - based intuitive interfaces, and automatic creation of structured queries and semi-structured data formats like yet another markup language (YAML), java script object notation (JSON), and application program interface (API) documentation. The system is intended to strengthen the potential of databases through the integration of Large Language Models (LLMs) and advanced machine learning algorithms. The integration is purposed to allow the automation of fundamental tasks such as data modeling, schema creation, query comprehension, and performance optimization. We present in this paper a system that aims to alleviate the main problems with current database technologies. It is meant to reduce the need for technical skills, manual tuning for better performance, and the potential for human error. The AI database employs generative schema inference and format selection to build its schema models and execution formats.