Parameter sharing, where each agent independently learns a policy with fully shared parameters between all policies, is a popular baseline method for multi-agent deep reinforcement learning. Unfortunately, since all agents share the same policy network, they cannot learn different policies or tasks. This issue has been circumvented experimentally by adding an agent-specific indicator signal to observations, which we term "agent indication". Agent indication is limited, however, in that without modification it does not allow parameter sharing to be applied to environments where the action spaces and/or observation spaces are heterogeneous. This work formalizes the notion of agent indication and proves that it enables convergence to optimal policies for the first time. Next, we formally introduce methods to extend parameter sharing to learning in heterogeneous observation and action spaces, and prove that these methods allow for convergence to optimal policies. Finally, we experimentally confirm that the methods we introduce function empirically, and conduct a wide array of experiments studying the empirical efficacy of many different agent indication schemes for image based observation spaces.
Since the introduction of DQN, a vast majority of reinforcement learning research has focused on reinforcement learning with deep neural networks as function approximators. New methods are typically evaluated on a set of environments that have now become standard, such as Atari 2600 games. While these benchmarks help standardize evaluation, their computational cost has the unfortunate side effect of widening the gap between those with ample access to computational resources, and those without. In this work we argue that, despite the community's emphasis on large-scale environments, the traditi…
Off-policy multi-step reinforcement learning algorithms consist of conservative and non-conservative algorithms: the former actively cut traces, whereas the latter do not. Recently, Munos et al. (2016) proved the convergence of conservative algorithms to an optimal Q-function. In contrast, non-conservative algorithms are thought to be unsafe and have a limited or no theoretical guarantee. Nonetheless, recent studies have shown that non-conservative algorithms empirically outperform conservative ones. Motivated by the empirical results and the lack of theory, we carry out theoretical analyses o…
Offline reinforcement learning enables agents to leverage large pre-collected datasets of environment transitions to learn control policies, circumventing the need for potentially expensive or unsafe online data collection. Significant progress has been made recently in offline model-based reinforcement learning, approaches which leverage a learned dynamics model. This typically involves constructing a probabilistic model, and using the model uncertainty to penalize rewards where there is insufficient data, solving for a pessimistic MDP that lower bounds the true MDP. Existing methods, however…
Many tasks in artificial intelligence require the collaboration of multiple agents. We exam deep reinforcement learning for multi-agent domains. Recent research efforts often take the form of two seemingly conflicting perspectives, the decentralized perspective, where each agent is supposed to have its own controller; and the centralized perspective, where one assumes there is a larger model controlling all agents. In this regard, we revisit the idea of the master-slave architecture by incorporating both perspectives within one framework. Such a hierarchical structure naturally leverages advan…
Several real-world scenarios, such as remote control and sensing, are comprised of action and observation delays. The presence of delays degrades the performance of reinforcement learning (RL) algorithms, often to such an extent that algorithms fail to learn anything substantial. This paper formally describes the notion of Markov Decision Processes (MDPs) with stochastic delays and shows that delayed MDPs can be transformed into equivalent standard MDPs (without delays) with significantly simplified cost structure. We employ this equivalence to derive a model-free Delay-Resolved RL framework a…
Revisiting Entropy Regularization: Adaptive Coefficient Unlocks Its Potential for LLM Reinforcement Learning
State Key Lab of Processors, Institute of Computing Technology, CAS（处理器国家重点实验室，计算技术研究所，中国科学院）；University of Science and Technology of China（中国科学技术大学）；University of Chinese Academy of Sciences（中国科学院大学）；StepFun Inc（StepFun公司）
National Yang Ming Chiao Tung University（国立阳明交通大学）；Facebook AI（脸书人工智能）；Academia Sinica（中央研究院）
Action-constrained reinforcement learning (RL) is a widely-used approach in various real-world applications, such as scheduling in networked systems with resource constraints and control of a robot with kinematic constraints. While the existing projection-based approaches ensure zero constraint violation, they could suffer from the zero-gradient problem due to the tight coupling of the policy gradient and the projection, which results in sample-inefficient training and slow convergence. To tackle this issue, we propose a learning algorithm that decouples the action constraints from the policy …
Recently, deep reinforcement learning (RL) algorithms have made great progress in multi-agent domain. However, due to characteristics of RL, training for complex tasks would be resource-intensive and time-consuming. To meet this challenge, mutual learning strategy between homogeneous agents is essential, which is under-explored in previous studies, because most existing methods do not consider to use the knowledge of agent models. In this paper, we present an adaptation method of the majority of multi-agent reinforcement learning (MARL) algorithms called KnowSR which takes advantage of the dif…
Southern University of Science and Technology（南方科技大学）；Research Institute of Trustworthy Autonomous System, SUSTech（南方科技大学可信自主系统研究院）
In this paper, we revisit endless online level generation with the recently proposed experience-driven procedural content generation via reinforcement learning (EDRL) framework. Inspired by an observation that EDRL tends to generate recurrent patterns, we formulate a notion of state space closure which makes any stochastic state appeared possibly in an infinite-horizon online generation process can be found within a finite-horizon. Through theoretical analysis, we find that even though state space closure arises a concern about diversity, it generalises EDRL trained with a finite-horizon to th…
Revisiting Generative Policies: A Simpler Reinforcement Learning Algorithmic Perspective
Shanghai Artificial Intelligence Laboratory（上海人工智能实验室）；Xi’an Jiaotong University（西安交通大学）；SenseTime Research（商汤科技研究院）；Peking University（北京大学）；The Chinese University of Hong Kong（香港中文大学）
Sharing parameters in multi-agent deep reinforcement learning has played an essential role in allowing algorithms to scale to a large number of agents. Parameter sharing between agents significantly decreases the number of trainable parameters, shortening training times to tractable levels, and has been linked to more efficient learning. However, having all agents share the same parameters can also have a detrimental effect on learning. We demonstrate the impact of parameter sharing methods on training speed and converged returns, establishing that when applied indiscriminately, their effectiv…
Actor-critic algorithms that make use of distributional policy evaluation have frequently been shown to outperform their non-distributional counterparts on many challenging control tasks. Examples of this behavior include the D4PG and DMPO algorithms as compared to DDPG and MPO, respectively [Barth-Maron et al., 2018; Hoffman et al., 2020]. However, both agents rely on the C51 critic for value estimation.One major drawback of the C51 approach is its requirement of prior knowledge about the minimum andmaximum values a policy can attain as well as the number of bins used, which fixes the resolut…
Handling the problem of scalability is one of the essential issues for multi-agent reinforcement learning (MARL) algorithms to be applied to real-world problems typically involving massively many agents. For this, parameter sharing across multiple agents has widely been used since it reduces the training time by decreasing the number of parameters and increasing the sample efficiency. However, using the same parameters across agents limits the representational capacity of the joint policy and consequently, the performance can be degraded in multi-agent tasks that require different behaviors fo…
The growth of autonomous vehicles, ridesharing systems, and self driving technology will bring a shift in the way ride hailing platforms plan out their services. However, these advances in technology coupled with road congestion, environmental concerns, fuel usage, vehicles emissions, and the high cost of the vehicle usage have brought more attention to better utilize the use of vehicles and their capacities. In this paper, we propose a novel multi-hop ride-sharing (MHRS) algorithm that uses deep reinforcement learning to learn optimal vehicle dispatch and matching decisions by interacting wit…
Revisiting Plasticity in Visual Reinforcement Learning: Data, Modules and Training Stages
Tsinghua University（清华大学）；The University of Sydney（悉尼大学）；Washington University in St. Louis（圣路易斯华盛顿大学）；JD Explore Academy（京东探索研究院）；Nanyang Technological University（南洋理工大学）
Revisiting Entropy in Reinforcement Learning for Large Reasoning Models
TJUNLP Lab, School of Computer Science and Technology, Tianjin University, China（天津大学计算机科学与技术学院 TJUNLP 实验室，中国）；Independent Researcher（独立研究者）；College of Computer and Information Engineering, Tianjin Normal University, China（天津师范大学计算机与信息工程学院，中国）
Achieving Collective Welfare in Multi-Agent Reinforcement Learning via Suggestion Sharing
Warwick Manufacturing Group, University of Warwick, Coventry, UK（沃里克制造集团，沃里克大学，科文特里，英国）；School of Electrical Engineering and Computer Science, Louisiana State University, USA（电气工程与计算机科学学院，路易斯安那州立大学，美国）；Department of Statistics, University of Warwick, Coventry, UK（统计学系，沃里克大学，科文特里，英国）；Alan Turing Institute, London, UK（艾伦·图灵研究所，伦敦，英国）
Shanghai Jiao Tong University（上海交通大学）；Tencent Inc.（腾讯公司）；The University of Manchester（曼彻斯特大学）
We revisit the estimation bias in policy gradients for the discounted episodic Markov decision process (MDP) from Deep Reinforcement Learning (DRL) perspective. The objective is formulated theoretically as the expected returns discounted over the time horizon. One of the major policy gradient biases is the state distribution shift: the state distribution used to estimate the gradients differs from the theoretical formulation in that it does not take into account the discount factor. Existing discussion of the influence of this bias was limited to the tabular and softmax cases in the literature…
Deep Reinforcement Learning (RL) is proven powerful for decision making in simulated environments. However, training deep RL model is challenging in real world applications such as production-scale health-care or recommender systems because of the expensiveness of interaction and limitation of budget at deployment. One aspect of the data inefficiency comes from the expensive hyper-parameter tuning when optimizing deep neural networks. We propose Adaptive Behavior Policy Sharing (ABPS), a data-efficient training algorithm that allows sharing of experience collected by behavior policy that is ad…
John A. Paulson School of Engineering and Applied Sciences, Harvard University（哈佛大学约翰·A·保尔森工程与应用科学学院）；Computer Science Department, Stanford University（斯坦福大学计算机科学系）；The Taub Faculty of Computer Science, Technion - Israel Institute of Technology（以色列理工学院陶布计算机科学学院）
Universitat Politècnica de Catalunya (UPC)（加泰罗尼亚理工大学）
The capacity sharing problem in Radio Access Network (RAN) slicing deals with the distribution of the capacity available in each RAN node among various RAN slices to satisfy their traffic demands and efficiently use the radio resources. While several capacity sharing algorithmic solutions have been proposed in the literature, their practical implementation still remains as a gap. In this paper, the implementation of a Reinforcement Learning-based capacity sharing algorithm over the O-RAN architecture is discussed, providing insights into the operation of the involved interfaces and the contain…
Princeton University（普林斯顿大学）；Carnegie Mellon University（卡内基梅隆大学）；Tsinghua University（清华大学）；University of Pennsylvania（宾夕法尼亚大学）
Low-complexity models such as linear function representation play a pivotal role in enabling sample-efficient reinforcement learning (RL). The current paper pertains to a scenario with value-based linear representation, which postulates the linear realizability of the optimal Q-function (also called the "linear $Q^{\star}$ problem"). While linear realizability alone does not allow for sample-efficient solutions in general, the presence of a large sub-optimality gap is a potential game changer, depending on the sampling mechanism in use. Informally, sample efficiency is achievable with a large …
Reinforcement learning (RL) has been used in a range of simulated real-world tasks, e.g., sensor coordination, traffic light control, and on-demand mobility services. However, real world deployments are rare, as RL struggles with dynamic nature of real world environments, requiring time for learning a task and adapting to changes in the environment. Transfer Learning (TL) can help lower these adaptation times. In particular, there is a significant potential of applying TL in multi-agent RL systems, where multiple agents can share knowledge with each other, as well as with new agents that join …
The University of Queensland（昆士兰大学）；Intuitive Surgical（直觉外科公司）
We provide new perspectives and inference algorithms for Maximum Entropy (MaxEnt) Inverse Reinforcement Learning (IRL), which provides a principled method to find a most non-committal reward function consistent with given expert demonstrations, among many consistent reward functions. We first present a generalized MaxEnt formulation based on minimizing a KL-divergence instead of maximizing an entropy. This improves the previous heuristic derivation of the MaxEnt IRL model (for stochastic MDPs), allows a unified view of MaxEnt IRL and Relative Entropy IRL, and leads to a model-free learning alg…
Numenta；Universidade do Vale do Rio dos Sinos（淡水河谷大学）；University of Brasilia（巴西利亚大学）
The idea of experience sharing between cooperative agents naturally emerges from our understanding of how humans learn. Our evolution as a species is tightly linked to the ability to exchange learned knowledge with one another. It follows that experience sharing (ES) between autonomous and independent agents could become the key to accelerate learning in cooperative multiagent settings. We investigate if randomly selecting experiences to share can increase the performance of deep reinforcement learning agents, and propose three new methods for selecting experiences to accelerate the learning p…
Tsinghua University（清华大学）；The Chinese University of Hong Kong（香港中文大学）
Bike sharing provides an environment-friendly way for traveling and is booming all over the world. Yet, due to the high similarity of user travel patterns, the bike imbalance problem constantly occurs, especially for dockless bike sharing systems, causing significant impact on service quality and company revenue. Thus, it has become a critical task for bike sharing systems to resolve such imbalance efficiently. In this paper, we propose a novel deep reinforcement learning framework for incentivizing users to rebalance such systems. We model the problem as a Markov decision process and take bot…
Learning Without Critics? Revisiting GRPO in Classical Reinforcement Learning Environments
Advanced Knowledge Center for Immersive Technologies（沉浸式技术高级知识中心）；NVIDIA；Institute of Informatics, Federal University of Goiás（戈亚斯联邦大学信息学院）；OATML, University of Oxford（牛津大学OATML）
Offline reinforcement learning (RL) algorithms have shown promising results in domains where abundant pre-collected data is available. However, prior methods focus on solving individual problems from scratch with an offline dataset without considering how an offline RL agent can acquire multiple skills. We argue that a natural use case of offline RL is in settings where we can pool large amounts of data collected in various scenarios for solving different tasks, and utilize all of this data to learn behaviors for all the tasks more effectively rather than training each one in isolation. Howeve…
Tsinghua University（清华大学）；Institute for Interdisciplinary Information Sciences, Tsinghua University（清华大学交叉信息研究院）
Self-supervised methods have become crucial for advancing deep learning by leveraging data itself to reduce the need for expensive annotations. However, the question of how to conduct self-supervised offline reinforcement learning (RL) in a principled way remains unclear. In this paper, we address this issue by investigating the theoretical benefits of utilizing reward-free data in linear Markov Decision Processes (MDPs) within a semi-supervised setting. Further, we propose a novel, Provable Data Sharing algorithm (PDS) to utilize such reward-free data for offline RL. PDS uses additional penal…
Centralized Reward Agent for Knowledge Sharing and Transfer in Multi-Task Reinforcement Learning
School of Computing, National University of Singapore（新加坡国立大学计算机学院）；TikTok Pte. Ltd., Singapore（新加坡TikTok公司）；School of Electrical and Electronic Engineering, Nanyang Technological University（南洋理工大学电子与电气工程学院）；Department of Electrical and Computer Engineering, National University of Singapore（新加坡国立大学电气与计算机工程系）
University of South Florida（南佛罗里达大学）；KAIST（韩国科学技术院）
This paper proposes a reinforcement learning approach for nightly offline rebalancing operations in free-floating electric vehicle sharing systems (FFEVSS). Due to sparse demand in a network, FFEVSS require relocation of electrical vehicles (EVs) to charging stations and demander nodes, which is typically done by a group of drivers. A shuttle is used to pick up and drop off drivers throughout the network. The objective of this study is to solve the shuttle routing problem to finish the rebalancing work in the minimal time. We consider a reinforcement learning framework for the problem, in whic…
International Institute of Information Technology, Hyderabad（海得拉巴国际信息技术学院）
Multi-agent learning provides a potential framework for learning and simulating traffic behaviors. This paper proposes a novel architecture to learn multiple driving behaviors in a traffic scenario. The proposed architecture can learn multiple behaviors independently as well as simultaneously. We take advantage of the homogeneity of agents and learn in a parameter sharing paradigm. To further speed up the training process asynchronous updates are employed into the architecture. While learning different behaviors simultaneously, the given framework was also able to learn cooperation between the…
Divisible Load Theory (DLT) is a powerful tool for modeling divisible load problems in data-intensive systems. This paper studied an optimal divisible load distribution sequencing problem using a machine learning framework. The problem is to decide the optimal sequence to distribute divisible load to processors in order to achieve minimum finishing time. The scheduling is performed in a resource-sharing system where each physical processor is virtualized to multiple virtual processors. A reinforcement learning method called Multi-armed bandit (MAB) is used for our problem. We first provide a n…
Significant development of ride-sharing services presents a plethora of opportunities to transform urban mobility by providing personalized and convenient transportation while ensuring efficiency of large-scale ride pooling. However, a core problem for such services is route planning for each driver to fulfill the dynamically arriving requests while satisfying given constraints. Current models are mostly limited to static routes with only two rides per vehicle (optimally) or three (with heuristics). In this paper, we present a dynamic, demand aware, and pricing-based vehicle-passenger matching…
Proximal Policy Optimization (PPO) is a popular deep policy gradient algorithm. In standard implementations, PPO regularizes policy updates with clipped probability ratios, and parameterizes policies with either continuous Gaussian distributions or discrete Softmax distributions. These design choices are widely accepted, and motivated by empirical performance comparisons on MuJoCo and Atari benchmarks. We revisit these practices outside the regime of current benchmarks, and expose three failure modes of standard PPO. We explain why standard design choices are problematic in these cases, and sh…
We have developed a model for online continual or lifelong reinforcement learning (RL) inspired on the insect brain. Our model leverages the offline training of a feature extraction and a common general policy layer to enable the convergence of RL algorithms in online settings. Sharing a common policy layer across tasks leads to positive backward transfer, where the agent continuously improved in older tasks sharing the same underlying general policy. Biologically inspired restrictions to the agent's network are key for the convergence of RL algorithms. This provides a pathway towards efficien…
Semantic-Aware Spectrum Sharing in Internet of Vehicles Based on Deep Reinforcement Learning
School of Internet of Things Engineering, Jiangnan University（江南大学物联网工程学院）；Department of Electronic Engineering, Beijing National Research Center for Information Science and Technology, Tsinghua University（清华大学电子工程系，北京信息科学与技术国家研究中心）；State Key Lab. of ISN and School of Telecommunications Engineering, Xidian University（西安电子科技大学信息与通信系统国家重点实验室及电信工程学院）；School of Engineering, University of Kent（肯特大学工程学院）；Department of Electrical and Computer Engineering, Hong Kong University of Science and Technology (HKUST)（香港科技大学电子与计算机工程系）
