ICML 2023 论文列表

We address multi-agent best arm identification with privacy guarantees. In this setting, agents collaborate by communicating to find the optimal arm. To avoid leaking sensitive data through messages, we consider two notions of privacy withholding different kinds of information: differential privacy and $(\epsilon, \eta)$-privacy. For each privacy definition, we propose an algorithm based on a two-level successive elimination scheme. We provide theoretical guarantees for the privacy level, accuracy and sample complexity of our algorithms. Experiments on various settings support our theoretical findings.

We consider the problem of multi-agent navigation and collision avoidance when observations are limited to the local neighborhood of each agent. We propose InforMARL, a novel architecture for multi-agent reinforcement learning (MARL) which uses local information intelligently to compute paths for all the agents in a decentralized manner. Specifically, InforMARL aggregates information about the local neighborhood of agents for both the actor and the critic using a graph neural network and can be used in conjunction with any standard MARL algorithm. We show that (1) in training, InforMARL has better sample efficiency and performance than baseline approaches, despite using less information, and (2) in testing, it scales well to environments with arbitrary numbers of agents and obstacles. We illustrate these results using four task environments, including one with predetermined goals for each agent, and one in which the agents collectively try to cover all goals.

A complex system with cluttered observations may be a coupled mixture of multiple simple sub-systems corresponding to latent entities. Such sub-systems may hold distinct dynamics in the continuous-time domain; therein, complicated interactions between sub-systems also evolve over time. This setting is fairly common in the real world but has been less considered. In this paper, we propose a sequential learning approach under this setting by decoupling a complex system for handling irregularly sampled and cluttered sequential observations. Such decoupling brings about not only subsystems describing the dynamics of each latent entity but also a meta-system capturing the interaction between entities over time. Specifically, we argue that the meta-system evolving within a simplex is governed by projected differential equations (ProjDEs). We further analyze and provide neural-friendly projection operators in the context of Bregman divergence. Experimental results on synthetic and real-world datasets show the advantages of our approach when facing complex and cluttered sequential data compared to the state-of-the-art.

The study of collaborative multi-agent bandits has attracted significant attention recently. In light of this, we initiate the study of a new collaborative setting, consisting of $N$ agents such that each agent is learning one of $M$ stochastic multi-armed bandits to minimize their group cumulative regret. We develop decentralized algorithms which facilitate collaboration between the agents under two scenarios. We characterize the performance of these algorithms by deriving the per agent cumulative regret and group regret upper bounds. We also prove lower bounds for the group regret in this setting, which demonstrates the near-optimal behavior of the proposed algorithms.

It has previously been reported that the representation that is learned in the first layer of deep Convolutional Neural Networks (CNNs) is highly consistent across initializations and architectures. In this work, we quantify this consistency by considering the first layer as a filter bank and measuring its energy distribution. We find that the energy distribution is very different from that of the initial weights and is remarkably consistent across random initializations, datasets, architectures and even when the CNNs are trained with *random labels*. In order to explain this consistency, we derive an analytical formula for the energy profile of linear CNNs and show that this profile is mostly dictated by the second order statistics of image patches in the training set and it will approach a whitening transformation when the number of iterations goes to infinity. Finally, we show that this formula for linear CNNs also gives an excellent fit for the energy profiles learned by commonly used *nonlinear* CNNs such as ResNet and VGG, and that the first layer of these CNNs indeed performs approximate whitening of their inputs.

Multi-step learning applies lookahead over multiple time steps and has proved valuable in policy evaluation settings. However, in the optimal control case, the impact of multi-step learning has been relatively limited despite a number of prior efforts. Fundamentally, this might be because multi-step policy improvements require operations that cannot be approximated by stochastic samples, hence hindering the widespread adoption of such methods in practice. To address such limitations, we introduce doubly multi-step off-policy VI (DoMo-VI), a novel oracle algorithm that combines multi-step policy improvements and policy evaluations. DoMo-VI enjoys guaranteed convergence speed-up to the optimal policy and is applicable in general off-policy learning settings. We then propose doubly multi-step off-policy actor-critic (DoMo-AC), a practical instantiation of the DoMo-VI algorithm. DoMo-AC introduces a bias-variance trade-off that ensures improved policy gradient estimates. When combined with the IMPALA architecture, DoMo-AC has showed improvements over the baseline algorithm on Atari-57 game benchmarks.

Ad hoc teamwork requires an agent to cooperate with unknown teammates without prior coordination. Many works propose to abstract teammate instances into high-level representation of types and then pre-train the best response for each type. However, most of them do not consider the distribution of teammate instances within a type. This could expose the agent to the hidden risk of type confounding. In the worst case, the best response for an abstract teammate type could be the worst response for all specific instances of that type. This work addresses the issue from the lens of causal inference. We first theoretically demonstrate that this phenomenon is due to the spurious correlation brought by uncontrolled teammate distribution. Then, we propose our solution, CTCAT, which disentangles such correlation through an instance-wise teammate feedback rectification. This operation reweights the interaction of teammate instances within a shared type to reduce the influence of type confounding. The effect of CTCAT is evaluated in multiple domains, including classic ad hoc teamwork tasks and real-world scenarios. Results show that CTCAT is robust to the influence of type confounding, a practical issue that directly hazards the robustness of our trained agents but was unnoticed in previous works.

Artificial neural networks are being proposed as models of parts of the brain. The networks are compared to recordings of biological neurons, and good performance in reproducing neural responses is considered to support the model's validity. A key question is how much this system identification approach tells us about brain computation. Does it validate one model architecture over another? We evaluate the most commonly used comparison techniques, such as a linear encoding model and centered kernel alignment, to correctly identify a model by replacing brain recordings with known ground truth models. System identification performance is quite variable; it also depends significantly on factors independent of the ground truth architecture, such as stimuli images. In addition, we show the limitations of using functional similarity scores in identifying higher-level architectural motifs.

In this paper, we propose an adaptive entropy-regularization framework (ADER) for multi-agent reinforcement learning (RL) to learn the adequate amount of exploration of each agent for entropy-based exploration. In order to derive a metric for the proper level of exploration entropy for each agent, we disentangle the soft value function into two types: one for pure return and the other for entropy. By applying multi-agent value factorization to the disentangled value function of pure return, we obtain a metric to determine the relevant level of exploration entropy for each agent, given by the partial derivative of the pure-return value function with respect to (w.r.t.) the policy entropy of each agent. Based on this metric, we propose the ADER algorithm based on maximum entropy RL, which controls the necessary level of exploration across agents over time by learning the proper target entropy for each agent. Experimental results show that the proposed scheme significantly outperforms current state-of-the-art multi-agent RL algorithms.

In order for agents in multi-agent systems (MAS) to be safe, they need to take into account the risks posed by the actions of other agents. However, the dominant paradigm in game theory (GT) assumes that agents are not affected by risk from other agents and only strive to maximise their expected utility. For example, in hybrid human-AI driving systems, it is necessary to limit large deviations in reward resulting from car crashes. Although there are equilibrium concepts in game theory that take into account risk aversion, they either assume that agents are risk-neutral with respect to the uncertainty caused by the actions of other agents, or they are not guaranteed to exist. We introduce a new GT-based Risk-Averse Equilibrium (RAE) that always produces a solution that minimises the potential variance in reward accounting for the strategy of other agents. Theoretically and empirically, we show RAE shares many properties with a Nash Equilibrium (NE), establishing convergence properties and generalising to risk-dominant NE in certain cases. To tackle large-scale problems, we extend RAE to the PSRO multi-agent reinforcement learning (MARL) framework. We empirically demonstrate the minimum reward variance benefits of RAE in matrix games with high-risk outcomes. Results on MARL experiments show RAE generalises to risk-dominant NE in a trust dilemma game and that it reduces instances of crashing by 7x in an autonomous driving setting versus the best performing baseline.

We study offline multi-agent reinforcement learning (RL) in Markov games, where the goal is to learn an approximate equilibrium---such as Nash equilibrium and (Coarse) Correlated Equilibrium---from an offline dataset pre-collected from the game. Existing works consider relatively restricted tabular or linear models and handle each equilibria separately. In this work, we provide the first framework for sample-efficient offline learning in Markov games under general function approximation, handling all 3 equilibria in a unified manner. By using Bellman-consistent pessimism, we obtain interval estimation for policies' returns, and use both the upper and the lower bounds to obtain a relaxation on the gap of a candidate policy, which becomes our optimization objective. Our results generalize prior works and provide several additional insights. Importantly, we require a data coverage condition that improves over the recently proposed ``unilateral concentrability''. Our condition allows selective coverage of deviation policies that optimally trade-off between their greediness (as approximate best responses) and coverage, and we show scenarios where this leads to significantly better guarantees. As a new connection, we also show how our algorithmic framework can subsume seemingly different solution concepts designed for the special case of two-player zero-sum games.

A large body of the "Inverse Reinforcement Learning" (IRL) literature focuses on recovering the reward function from a set of demonstrations of an expert agent who acts optimally or noisily optimally. Nevertheless, some recent works move away from the optimality assumption to study the "Learning from a Learner (LfL)" problem, where the challenge is inferring the reward function of a learning agent from a sequence of demonstrations produced by progressively improving policies. In this work, we take one of the initial steps in addressing the multi-agent version of this problem and propose a new algorithm, MA-LfL (Multiagent Learning from a Learner). Unlike the state-of-the-art literature, which recovers the reward functions from trajectories produced by agents in some equilibrium, we study the problem of inferring the reward functions of interacting agents in a general sum stochastic game without assuming any equilibrium state. The MA-LfL algorithm is rigorously built on a theoretical result that ensures its validity in the case of agents learning according to a multi-agent soft policy iteration scheme. We empirically test MA-LfL and we observe high positive correlation between the recovered reward functions and the ground truth.

Sparse reward remains a valuable and challenging problem in multi-agent reinforcement learning (MARL). This paper addresses this issue from a new perspective, i.e., lazy agents. We empirically illustrate how lazy agents damage learning from both exploration and exploitation. Then, we propose a novel MARL framework called Lazy Agents Avoidance through Influencing External States (LAIES). Firstly, we examine the causes and types of lazy agents in MARL using a causal graph of the interaction between agents and their environment. Then, we mathematically define the concept of fully lazy agents and teams by calculating the causal effect of their actions on external states using the do-calculus process. Based on definitions, we provide two intrinsic rewards to motivate agents, i.e., individual diligence intrinsic motivation (IDI) and collaborative diligence intrinsic motivation (CDI). IDI and CDI employ counterfactual reasoning based on the external states transition model (ESTM) we developed. Empirical results demonstrate that our proposed method achieves state-of-the-art performance on various tasks, including the sparse-reward version of StarCraft multi-agent challenge (SMAC) and Google Research Football (GRF). Our code is open-source and available at https://github.com/liuboyin/LAIES.

We investigate the problem of best arm identification in Multi-Agent Multi-Armed Bandits (MAMABs) where the rewards are defined through a factor graph. The objective is to find an optimal global action with a prescribed level of confidence and minimal sample complexity. We derive a tight instance-specific lower bound of the sample complexity and characterize the corresponding optimal sampling strategy. Unfortunately, this bound is obtained by solving a combinatorial optimization problem with a number of variables and constraints exponentially growing with the number of agents. We leverage Mean Field (MF) techniques to obtain, in a computationally efficient manner, an approximation of the lower bound. The approximation scales at most as $\rho K^d$ (where $\rho$, $K$, and $d$ denote the number of factors in the graph, the number of possible actions per agent, and the maximal degree of the factor graph). We devise MF-TaS (Mean-Field-Track-and-Stop), an algorithm whose sample complexity provably matches our approximated lower bound. We illustrate the performance of MF-TaS numerically using both synthetic and real-world experiments (e.g., to solve the antenna tilt optimization problem in radio communication networks).

An unaddressed challenge in multi-agent coordination is to enable AI agents to exploit the semantic relationships between the features of actions and the features of observations. Humans take advantage of these relationships in highly intuitive ways. For instance, in the absence of a shared language, we might point to the object we desire or hold up our fingers to indicate how many objects we want. To address this challenge, we investigate the effect of network architecture on the propensity of learning algorithms to exploit these semantic relationships. Across a procedurally generated coordination task, we find that attention-based architectures that jointly process a featurized representation of observations and actions have a better inductive bias for learning intuitive policies. Through fine-grained evaluation and scenario analysis, we show that the resulting policies are human-interpretable. Moreover, such agents coordinate with people without training on any human data.

This paper considers the unlabeled sparse recovery under multiple measurements, i.e., ${\mathbf{Y}} = {\mathbf{\Pi}}^{\natural} {\mathbf{X}} {\mathbf{B}}^{\natural} + {\mathbf{W}}$, where ${\mathbf{Y}} \in \mathbb{R}^{n\times m}, {\mathbf{\Pi}}^{\natural}\in \mathbb{R}^{n\times n}, {\mathbf{X}} \in \mathbb{R}^{n\times p}, {\mathbf{B}} ^{\natural}\in \mathbb{R}^{p\times m}, {\mathbf{W}}\in \mathbb{R}^{n\times m}$ represents the observations, missing (or incomplete) correspondence information, sensing matrix, sparse signals, and additive sensing noise, respectively. Different from the previous works on multiple measurements ($m > 1$) which all focus on the sufficient samples regime, namely, $n > p$, we consider a sparse matrix $\mathbf{B}^{\natural}$ and investigate the insufficient samples regime (i.e., $n \ll p$) for the first time. To begin with, we establish the lower bound on the sample number and *signal-to-noise ratio* ($ {\mathsf{SNR}}$) for the correct permutation recovery. Moreover, we present a simple yet effective estimator. Under mild conditions, we show that our estimator can restore the correct correspondence information with high probability. Numerical experiments are presented to corroborate our theoretical claims.

Large transformer models powered by diverse data and model scale have dominated natural language modeling and computer vision and pushed the frontier of multiple AI areas. In reinforcement learning (RL), despite many efforts into transformer-based policies, a key limitation, however, is that current transformer-based policies cannot learn by directly combining information from multiple sub-optimal trials. In this work, we address this issue using recently proposed chain of hindsight to relabel experience, where we train a transformer on a sequence of trajectory experience ascending sorted according to their total rewards. Our method consists of relabelling target return of each trajectory to the maximum total reward among in sequence of trajectories and training an autoregressive model to predict actions conditioning on past states, actions, rewards, target returns, and task completion tokens, the resulting model, Agentic Transformer (AT), can learn to improve upon itself both at training and test time. As we show on D4RL and ExoRL benchmarks, to the best our knowledge, this is the first time that a simple transformer-based model performs competitively with both temporal-difference and imitation-learning-based approaches, even from sub-optimal data. Our Agentic Transformer also shows a promising scaling trend that bigger models consistently improve results.

Learning agent behaviors from observational data has shown to improve our understanding of their decision-making processes, advancing our ability to explain their interactions with the environment and other agents. While multiple learning techniques have been proposed in the literature, there is one particular setting that has not been explored yet: multi agent systems where agent identities remain anonymous. For instance, in financial markets labeled data that identifies market participant strategies is typically proprietary, and only the anonymous state-action pairs that result from the interaction of multiple market participants are publicly available. As a result, sequences of agent actions are not observable, restricting the applicability of existing work. In this paper, we propose a Policy Clustering algorithm, called K-SHAP, that learns to group anonymous state-action pairs according to the agent policies. We frame the problem as an Imitation Learning (IL) task, and we learn a world-policy able to mimic all the agent behaviors upon different environmental states. We leverage the world-policy to explain each anonymous observation through an additive feature attribution method called SHAP (SHapley Additive exPlanations). Finally, by clustering the explanations we show that we are able to identify different agent policies and group observations accordingly. We evaluate our approach on simulated synthetic market data and a real-world financial dataset. We show that our proposal significantly and consistently outperforms the existing methods, identifying different agent strategies.