Matrix games like Prisoner's Dilemma have guided research on social dilemmas for decades. However, they necessarily treat the choice to cooperate or defect as an atomic action. In real-world social dilemmas these choices are temporally extended. Cooperativeness is a property that applies to policies, not elementary actions. We introduce sequential social dilemmas that share the mixed incentive structure of matrix game social dilemmas but also require agents to learn policies that implement their strategic intentions. We analyze the dynamics of policies learned by multiple self-interested independent learning agents, each using its own deep Q-network, on two Markov games we introduce here: 1. a fruit Gathering game and 2. a Wolfpack hunting game. We characterize how learned behavior in each domain changes as a function of environmental factors including resource abundance. Our experiments show how conflict can emerge from competition over shared resources and shed light on how the sequential nature of real world social dilemmas affects cooperation.
Media and Data Science Research Lab, Adobe（Adobe媒体与数据科学研究实验室）；IIT Kharagpur, India（印度理工学院卡格拉普尔分校）
In social dilemma situations, individual rationality leads to sub-optimal group outcomes. Several human engagements can be modeled as a sequential (multi-step) social dilemmas. However, in contrast to humans, Deep Reinforcement Learning agents trained to optimize individual rewards in sequential social dilemmas converge to selfish, mutually harmful behavior. We introduce a status-quo loss (SQLoss) that encourages an agent to stick to the status quo, rather than repeatedly changing its policy. We show how agents trained with SQLoss evolve cooperative behavior in several social dilemma matrix ga…
Tianjin University（天津大学）；Washington State University（华盛顿州立大学）
The Iterated Prisoner's Dilemma has guided research on social dilemmas for decades. However, it distinguishes between only two atomic actions: cooperate and defect. In real-world prisoner's dilemmas, these choices are temporally extended and different strategies may correspond to sequences of actions, reflecting grades of cooperation. We introduce a Sequential Prisoner's Dilemma (SPD) game to better capture the aforementioned characteristics. In this work, we propose a deep multiagent reinforcement learning approach that investigates the evolution of mutual cooperation in SPD games. Our approa…
Groups of humans are often able to find ways to cooperate with one another in complex, temporally extended social dilemmas. Models based on behavioral economics are only able to explain this phenomenon for unrealistic stateless matrix games. Recently, multi-agent reinforcement learning has been applied to generalize social dilemma problems to temporally and spatially extended Markov games. However, this has not yet generated an agent that learns to cooperate in social dilemmas as humans do. A key insight is that many, but not all, human individuals have inequity averse social preferences. This…
Reciprocity is an important feature of human social interaction and underpins our cooperative nature. What is more, simple forms of reciprocity have proved remarkably resilient in matrix game social dilemmas. Most famously, the tit-for-tat strategy performs very well in tournaments of Prisoner's Dilemma. Unfortunately this strategy is not readily applicable to the real world, in which options to cooperate or defect are temporally and spatially extended. Here, we present a general online reinforcement learning algorithm that displays reciprocal behavior towards its co-players. We show that it c…
Multi-Agent, Human-Agent and Beyond: A Survey on Cooperation in Social Dilemmas
Northwestern Polytechnical University（西北工业大学）；Tsinghua University（清华大学）；University of Auckland（奥克兰大学）；Kyushu University（九州大学）；The Hong Kong Polytechnic University（香港理工大学）；Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）
The Alan Turing Institute（艾伦图灵研究所）；University College London（伦敦大学学院）
Multi-agent reinforcement learning has received significant interest in recent years notably due to the advancements made in deep reinforcement learning which have allowed for the developments of new architectures and learning algorithms. Using social dilemmas as the training ground, we present a novel learning architecture, Learning through Probing (LTP), where agents utilize a probing mechanism to incorporate how their opponent's behavior changes when an agent takes an action. We use distinct training phases and adjust rewards according to the overall outcome of the experiences accounting fo…
Multi-agent reinforcement learning (MARL) has shown recent success in increasingly complex fixed-team zero-sum environments. However, the real world is not zero-sum nor does it have fixed teams; humans face numerous social dilemmas and must learn when to cooperate and when to compete. To successfully deploy agents into the human world, it may be important that they be able to understand and help in our conflicts. Unfortunately, selfish MARL agents typically fail when faced with social dilemmas. In this work, we show evidence of emergent direct reciprocity, indirect reciprocity and reputation, …
When an individual's behavior has rational characteristics, this may lead to irrational collective actions for the group. A wide range of organisms from animals to humans often evolve the social attribute of cooperation to meet this challenge. Therefore, cooperation among individuals is of great significance for allowing social organisms to adapt to changes in the natural environment. Based on multi-agent reinforcement learning, we propose a new learning strategy for achieving coordination by incorporating a learning rate that can balance exploration and exploitation. We demonstrate that agent…
Achieving and maintaining cooperation between agents to accomplish a common objective is one of the central goals of Multi-Agent Reinforcement Learning (MARL). Nevertheless in many real-world scenarios, separately trained and specialized agents are deployed into a shared environment, or the environment requires multiple objectives to be achieved by different coexisting parties. These variations among specialties and objectives are likely to cause mixed motives that eventually result in a social dilemma where all the parties are at a loss. In order to resolve this issue, we propose the Incentiv…
The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））；East China Normal University（华东师范大学）
Social dilemmas can be considered situations where individual rationality leads to collective irrationality. The multi-agent reinforcement learning community has leveraged ideas from social science, such as social value orientations (SVO), to solve social dilemmas in complex cooperative tasks. In this paper, by first introducing the typical "division of labor or roles" mechanism in human society, we provide a promising solution for intertemporal social dilemmas (ISD) with SVOs. A novel learning framework, called Learning Roles with Emergent SVOs (RESVO), is proposed to transform the learning o…
Fairness over Equality: Correcting Social Incentives in Asymmetric Sequential Social Dilemmas
Izmir University of Economics \& University of Edinburgh（伊兹密大学经济学院及爱丁堡大学）；Middle East Technical University \& Utrecht University（中东部技术大学及乌得勒支大学）；University of Edinburgh（爱丁堡大学）；DeepFlow
As an important psychological and social experiment, the Iterated Prisoner's Dilemma (IPD) treats the choice to cooperate or defect as an atomic action. We propose to study the behaviors of online learning algorithms in the Iterated Prisoner's Dilemma (IPD) game, where we investigate the full spectrum of reinforcement learning agents: multi-armed bandits, contextual bandits and reinforcement learning. We evaluate them based on a tournament of iterated prisoner's dilemma where multiple agents can compete in a sequential fashion. This allows us to analyze the dynamics of policies learned by mult…
Intention Inference Under Execution Noise: Separating Aleatoric and Epistemic Uncertainty in Social Dilemmas
University of Cape Town（开普敦大学）；Neuroscience Institute, University of Cape Town（开普敦大学神经科学研究所）；National Institute for Theoretical & Computational Sciences, Stellenbosch（斯泰伦博斯理论与计算科学国家研究所）；Institut National de la recherche scientifique, Montreal（魁北克国家科学研究院（蒙特利尔））
Institution for Interdisciplinary Information Sciences（交叉信息研究院）；Tsinghua University（清华大学）
How cooperation emerges is a long-standing and interdisciplinary problem. Game-theoretical studies on social dilemmas reveal that altruistic incentives are critical to the emergence of cooperation but their analyses are limited to stateless games. For more realistic scenarios, multi-agent reinforcement learning has been used to study sequential social dilemmas (SSDs). Recent works show that learning to incentivize other agents can promote cooperation in SSDs. However, we find that, with these incentivizing mechanisms, the team cooperation level does not converge and regularly oscillates betwee…
Social dilemmas are situations where individuals face a temptation to increase their payoffs at a cost to total welfare. Building artificially intelligent agents that achieve good outcomes in these situations is important because many real world interactions include a tension between selfish interests and the welfare of others. We show how to modify modern reinforcement learning methods to construct agents that act in ways that are simple to understand, nice (begin by cooperating), provokable (try to avoid being exploited), and forgiving (try to return to mutual cooperation). We show both theo…
Social dilemmas, where mutual cooperation can lead to high payoffs but participants face incentives to cheat, are ubiquitous in multi-agent interaction. We wish to construct agents that cooperate with pure cooperators, avoid exploitation by pure defectors, and incentivize cooperation from the rest. However, often the actions taken by a partner are (partially) unobserved or the consequences of individual actions are hard to predict. We show that in a large class of games good strategies can be constructed by conditioning one's behavior solely on outcomes (ie. one's past rewards). We call this c…
It has been argued that one role of social constructs, such as institutions, trust and norms, is to coordinate the expectations of autonomous entities in order to resolve collective action situations (such as collective risk dilemmas) through the coordination of behaviour. While much work has addressed the formal representation of these social constructs, in this paper we focus specifically on the formal representation of, and associated reasoning with, the expectations themselves. In particular, we investigate how explicit reasoning about expectations can be used to encode both traditional ga…
When developing reinforcement learning agents, the standard approach is to train an agent to converge to a fixed policy that is as close to optimal as possible for a single fixed reward function. If different agent behaviour is required in the future, an agent trained in this way must normally be either fully or partially retrained, wasting valuable time and resources. In this study, we leverage multi-objective reinforcement learning to create tunable agents, i.e. agents that can adopt a range of different behaviours according to the designer's preferences, without the need for retraining. We …
LOPT: Learning Optimal Pigovian Tax in Sequential Social Dilemmas
Antai College of Economics and Management, Shanghai Jiao Tong University（上海交通大学安泰经济与管理学院）；School of Computer Science and Technology, East China Normal University（华东师范大学计算机科学与技术学院）；School of Computer Science and Technology, Tongji University（同济大学计算机科学与技术学院）；Key Laboratory of Mathematics and Engineering Applications (MoE)（数学与工程应用教育部重点实验室）；Shanghai Institute of AI for Education, East China Normal University（华东师范大学上海人工智能教育研究院）；Shenzhen Loop Area Institute (SLAI)（深圳环路研究院）；School of Data Science, Chinese University of Hong Kong (Shenzhen)（香港中文大学（深圳）数据科学学院）
CoopEval: Benchmarking Cooperation-Sustaining Mechanisms and LLM Agents in Social Dilemmas
Carnegie Mellon University；Foundations of Cooperative AI Lab (FOCAL)；Jinesis Lab, University of Toronto \& Vector Institute；ETH Z\" u rich；Max Planck Institute for Intelligent Systems, T\" u bingen, Germany
Modeling Decision-Making with Will for Cooperation in Social Dilemmas
School of Intelligence Science and Technology, Peking University（北京大学智能学院）；State Key Laboratory of General Artificial Intelligence, BIGAI（通用人工智能国家重点实验室，BIGAI）；Law School, Peking University（北京大学法学院）
Situations in which immediate self-interest and long-term collective interest conflict often require some form of influence to prevent them from leading to undesirable or unsustainable outcomes. Next to sanctioning, social influence and social structure, it is possible that strategic solutions can exist for these social dilemmas. However, the existence of strategies that enable a player to exert control in the long-run outcomes can be difficult to show and different situations allow for different levels of strategic influence. Here, we investigate the effect of threshold nonlinearities on the …
We present our preliminary work on a multi-agent system involving the complex human phenomena of identity and dynamic teams. We outline our ongoing experimentation into understanding how these factors can eliminate some of the naive assumptions of current multi-agent approaches. These include a lack of complex heterogeneity between agents and unchanging team structures. We outline the human social psychological basis for identity, one's sense of self, and dynamic teams, the changing nature of human teams. We describe our application of these factors to a multi-agent system and our expectations…
In recent years, Win-Stay-Lose-Learn rule has attracted wide attention as an effective strategy updating rule, and voluntary participation is proposed by introducing a third strategy in Prisoner's dilemma game. Some researches show that combining Win-Stay-Lose-Learn rule with voluntary participation could promote cooperation more significantly under moderate temptation values, however, cooperators' survival under high aspiration levels and high temptation values is still a challenging problem. In this paper, inspired by Achievement Motivation Theory, a Dynamic-Win-Stay-Lose-Learn rule with vol…
Interactions among individuals in natural populations often occur in a dynamically changing environment. Understanding the role of environmental variation in population dynamics has long been a central topic in theoretical ecology and population biology. However, the key question of how individuals, in the middle of challenging social dilemmas (e.g., the "tragedy of the commons"), modulate their behaviors to adapt to the fluctuation of the environment has not yet been addressed satisfactorily. Utilizing evolutionary game theory and stochastic games, we develop a game-theoretical framework that…
Individual rationality, which involves maximizing expected individual returns, does not always lead to high-utility individual or group outcomes in multi-agent problems. For instance, in multi-agent social dilemmas, Reinforcement Learning (RL) agents trained to maximize individual rewards converge to a low-utility mutually harmful equilibrium. In contrast, humans evolve useful strategies in such social dilemmas. Inspired by ideas from human psychology that attribute this behavior to the status-quo bias, we present a status-quo loss (SQLoss) and the corresponding policy gradient algorithm that …
