arXiv subjects

Luis Felipe Giraldo

Publications and source records attributed to Luis Felipe Giraldo.

10 recordsLinked to original sources

arXiv2026

Learning Incentive Structures for Cooperative Resilience in Multi-Agent Systems under Social Dilemmas

Multi-agent social dilemmas, such as the tragedy of the commons, capture settings where individual incentives conflict with collective well-being, making these systems highly vulnerable to collapse under disruptions. In this context, this work studies cooperative resilience, understood as the system-level ability to maintain collective well-being under perturbations through adaptive agent behavior. We propose a framework for learning incentive structures aligned with collective well-being in multi-agent reinforcement learning systems, where reward functions shape individual decision-making and collective behavior. A resilience metric is used to score and rank agent trajectories, allowing the inference of reward functions that promote resilient collective behavior. These inferred reward functions are integrated into the multi-agent reinforcement learning process to shape agent interactions in social dilemma settings. The approach is evaluated in resource-sharing environments subject to disruptions, using three incentive structures: individual incentives, resilience-aligned incentives, and a hybrid incentive structure that combines both individual and collective components. The results show that the hybrid incentive structure promotes sustained collective behavior, reduces collapse events associated with resource depletion, and preserves system performance under disruption. These findings highlight the role of incentive design as a mechanism for promoting resilient collective behavior and provide a computational framework for multi-agent social dilemmas under disruptions.

cs.MA↗

arXiv2025

Evaluating Cooperative Resilience in Multiagent Systems: A Comparison Between Humans and LLMs

This paper presents a comparative analysis of cooperative resilience in multi-agent systems, defined as the ability to anticipate, resist, recover from, and transform to disruptive events that affect collective well-being. We focus on mixed-motive social dilemmas instantiated as a \textit{Tragedy of the Commons} environment from the Melting Pot suite, where we systematically compare human groups and Large Language Model (LLM)-based agents, each evaluated with and without explicit communication. Cooperative resilience is assessed under a continuously disruptive condition induced by a persistent unsustainable consumption bot, together with intermittent environmental shocks implemented as stochastic removal of shared resources across scenarios. This experimental design establishes a benchmark for cooperative resilience across agent architectures and interaction modalities, constituting a key step toward systematically comparing humans and LLM-based agents. Using this framework, we find that human groups with communication achieve the highest cooperative resilience compared to all other groups. Communication also improves the resilience of LLM agents, but their performance remains below human levels. Motivated by the performance of humans, we further examine a long-horizon setting with harsher environmental conditions, where humans sustain the shared resource and maintain high resilience in diverse disruption scenarios. Together, these results suggest that human decision-making under adverse social conditions can inform the design of artificial agents that promote prosocial and resilient behaviors.

cs.MA↗

arXiv2025

Evaluating Generalization Capabilities of LLM-Based Agents in Mixed-Motive Scenarios Using Concordia

Large Language Model (LLM) agents have demonstrated impressive capabilities for social interaction and are increasingly being deployed in situations where they might engage with both human and artificial agents. These interactions represent a critical frontier for LLM-based agents, yet existing evaluation methods fail to measure how well these capabilities generalize to novel social situations. In this paper, we introduce a method for evaluating the ability of LLM-based agents to cooperate in zero-shot, mixed-motive environments using Concordia, a natural language multi-agent simulation environment. Our method measures general cooperative intelligence by testing an agent's ability to identify and exploit opportunities for mutual gain across diverse partners and contexts. We present empirical results from the NeurIPS 2024 Concordia Contest, where agents were evaluated on their ability to achieve mutual gains across a suite of diverse scenarios ranging from negotiation to collective action problems. Our findings reveal significant gaps between current agent capabilities and the robust generalization required for reliable cooperation, particularly in scenarios demanding persuasion and norm enforcement.

cs.AI↗

arXiv2024

Cooperative Resilience in Artificial Intelligence Multiagent Systems

Resilience refers to the ability of systems to withstand, adapt to, and recover from disruptive events. While studies on resilience have attracted significant attention across various research domains, the precise definition of this concept within the field of cooperative artificial intelligence remains unclear. This paper addresses this gap by proposing a clear definition of `cooperative resilience' and outlining a methodology for its quantitative measurement. The methodology is validated in an environment with RL-based and LLM-augmented autonomous agents, subjected to environmental changes and the introduction of agents with unsustainable behaviors. These events are parameterized to create various scenarios for measuring cooperative resilience. The results highlight the crucial role of resilience metrics in analyzing how the collective system prepares for, resists, recovers from, sustains well-being, and transforms in the face of disruptions. These findings provide foundational insights into the definition, measurement, and preliminary analysis of cooperative resilience, offering significant implications for the broader field of AI. Moreover, the methodology and metrics developed here can be adapted to a wide range of AI applications, enhancing the reliability and effectiveness of AI in dynamic and unpredictable environments.

cs.MA↗

arXiv2024

OGLE-IV Period-Luminosity relation of the LMC: an analysis using mean and median magnitudes

TThe Period-Luminosity (PL) relation for Cepheid variable stars in the Large Magellanic Cloud (LMC) is crucial for distance measurements in astronomy. This study analyzes the impact of using the median rather than the mean on the PL relation's slope and zero point. It also examines the persistence of the break at approximately 10 days and addresses specification issues in the PL relation model. Using $VI$-band median and mean magnitudes from the OGLE-IV survey, corrected for extinction, we fit the PL relation employing robust $MM$-regression, which features a high breakdown point and robust standard errors. Statistical tests and residual analysis are conducted to identify and correct model deficiencies. Our findings indicate a significant change in the PL relation for Cepheids with periods of 10 days or longer, regardless of whether median or mean magnitudes are used. A bias in the zero point and slope estimators is observed when using median magnitudes instead of mean magnitudes, especially in the $V$-band. By identifying and correcting regression issues and considering the period break, our estimators for slope and zero point are more accurate for distance calculations. Comparative analysis of the models for each band quantifies the bias introduced by using median magnitudes, highlighting the importance of considering the Cepheids' period for accurate location measure results, similar to those obtained using mean magnitudes.

astro-ph.SR↗

arXiv2024

Can LLM-Augmented autonomous agents cooperate?, An evaluation of their cooperative capabilities through Melting Pot

As the field of AI continues to evolve, a significant dimension of this progression is the development of Large Language Models and their potential to enhance multi-agent artificial intelligence systems. This paper explores the cooperative capabilities of Large Language Model-augmented Autonomous Agents (LAAs) using the well-known Meltin Pot environments along with reference models such as GPT4 and GPT3.5. Preliminary results suggest that while these agents demonstrate a propensity for cooperation, they still struggle with effective collaboration in given environments, emphasizing the need for more robust architectures. The study's contributions include an abstraction layer to adapt Melting Pot game scenarios for LLMs, the implementation of a reusable architecture for LLM-mediated agent development - which includes short and long-term memories and different cognitive modules, and the evaluation of cooperation capabilities using a set of metrics tied to the Melting Pot's "Commons Harvest" game. The paper closes, by discussing the limitations of the current architectural framework and the potential of a new set of modules that fosters better cooperation among LAAs.

cs.AI↗

arXiv2023

Understanding the World to Solve Social Dilemmas Using Multi-Agent Reinforcement Learning

Social dilemmas are situations where groups of individuals can benefit from mutual cooperation but conflicting interests impede them from doing so. This type of situations resembles many of humanity's most critical challenges, and discovering mechanisms that facilitate the emergence of cooperative behaviors is still an open problem. In this paper, we study the behavior of self-interested rational agents that learn world models in a multi-agent reinforcement learning (RL) setting and that coexist in environments where social dilemmas can arise. Our simulation results show that groups of agents endowed with world models outperform all the other tested ones when dealing with scenarios where social dilemmas can arise. We exploit the world model architecture to qualitatively assess the learnt dynamics and confirm that each agent's world model is capable to encode information of the behavior of the changing environment and the other agent's actions. This is the first work that shows that world models facilitate the emergence of complex coordinated behaviors that enable interacting agents to ``understand'' both environmental and social dynamics.

cs.LG↗

arXiv2023

Climate change impact on photovoltaic power potential in South America

This paper presents the first study of the long-term impact of climate change on photovoltaic potential in South America. This region has great potential for implementing renewable energy, mainly solar energy solutions, due to its high solar irradiance levels. Based on the Coordinated Regional Downscaling Experiment (CORDEX) for the South American region, we estimate how climate change could affect photovoltaic power potential by the end of the century. The evidence suggests that photovoltaic potential could have a maximum decrease of $15\%$, and a maximum increase of $7\%$, primarily due to changes in solar irradiance of different zones. Furthermore, it is observed that regions with increased temperature also show increased solar irradiance levels, which could, to some extent, compensate for the losses caused by the rise in temperature. Therefore, photovoltaic production in most of the territory will not be negatively affected.

physics.ao-ph↗

arXiv2022

Learning Transferable Concepts in Deep Reinforcement Learning

While humans and animals learn incrementally during their lifetimes and exploit their experience to solve new tasks, standard deep reinforcement learning methods specialize to solve only one task at a time. As a result, the information they acquire is hardly reusable in new situations. Here, we introduce a new perspective on the problem of leveraging prior knowledge to solve future tasks. We show that learning discrete representations of sensory inputs can provide a high-level abstraction that is common across multiple tasks, thus facilitating the transference of information. In particular, we show that it is possible to learn such representations by self-supervision, following an information theoretic approach. Our method is able to learn concepts in locomotive and optimal control tasks that increase the sample efficiency in both known and unknown tasks, opening a new path to endow artificial agents with generalization abilities.

cs.AI↗

arXiv2020

Data-driven Outer-Loop Control Using Deep Reinforcement Learning for Trajectory Tracking

Reference tracking systems involve a plant that is stabilized by a local feedback controller and a command center that indicates the reference set-point the plant should follow. Typically, these systems are subject to limitations such as disturbances, systems delays, constraints, uncertainties, underperforming controllers, and unmodeled parameters that do not allow them to achieve the desired performance. In situations where it is not possible to redesign the inner-loop system, it is usual to incorporate an outer-loop control that instructs the system to follow a modified reference path such that the resultant path is close to the ideal one. Typically, strategies to design the outer-loop control need to know a model of the system, which can be an unfeasible task. In this paper, we propose a framework based on deep reinforcement learning that can learn a policy to generate a modified reference that improves the system's performance in a non-invasive and model-free fashion. To illustrate the effectiveness of our approach, we present two challenging cases in engineering: a flight control with a pilot model that includes human reaction delays, and a mean-field control problem for a massive number of space-heating devices. The proposed strategy successfully designs a reference signal that works even in situations that were not seen during the learning process.

eess.SY↗