NeurIPS 2023 Papers

The $k$-means++ algorithm of Arthur and Vassilvitskii (SODA 2007) is often the practitioners' choice algorithm for optimizing the popular $k$-means clustering objective and is known to give an $O(\log k)$-approximation in expectation. To obtain higher quality solutions, Lattanzi and Sohler (ICML 2019) proposed augmenting $k$-means++ with $O(k \log \log k)$ local-search steps obtained through the $k$-means++ sampling distribution to yield a $c$-approximation to the $k$-means clustering problem, where $c$ is a large absolute constant. Here we generalize and extend their local-search algorithm by considering larger and more sophisticated local-search neighborhoods hence allowing to swap multiple centers at the same time. Our algorithm achieves a $9 + \varepsilon$ approximation ratio, which is the best possible for local search. Importantly we show that our algorithm is practical, namely easy to implement and fast enough to run on a variety of classic datasets, and outputs solutions of better cost.

Instance detection (InsDet) is a long-lasting problem in robotics and computer vision, aiming to detect object instances (predefined by some visual examples) in a cluttered scene. Despite its practical significance, its advancement is overshadowed by Object Detection, which aims to detect objects belonging to some predefined classes. One major reason is that current InsDet datasets are too small in scale by today's standards. For example, the popular InsDet dataset GMU (published in 2016) has only 23 instances, far less than COCO (80 classes), a well-known object detection dataset published in 2014. We are motivated to introduce a new InsDet dataset and protocol. First, we define a realistic setup for InsDet: training data consists of multi-view instance captures, along with diverse scene images allowing synthesizing training images by pasting instance images on them with free box annotations. Second, we release a real-world database, which contains multi-view capture of 100 object instances, and high-resolution (6k$\times$8k) testing images. Third, we extensively study baseline methods for InsDet on our dataset, analyze their performance and suggest future work. Somewhat surprisingly, using the off-the-shelf class-agnostic segmentation model (Segment Anything Model, SAM) and the self-supervised feature representation DINOv2 performs the best, achieving $>$10 AP better than end-to-end trained InsDet models that repurpose object detectors (e.g., FasterRCNN and RetinaNet).

We provide a unifying framework for the design and analysis of multi-calibrated predictors. By placing the multi-calibration problem in the general setting of multi-objective learning---where learning guarantees must hold simultaneously over a set of distributions and loss functions---we exploit connections to game dynamics to achieve state-of-the-art guarantees for a diverse set of multi-calibration learning problems. In addition to shedding light on existing multi-calibration guarantees and greatly simplifying their analysis, our approach also yields improved guarantees, such as error tolerances that scale with the square-root of group size versus the constant tolerances guaranteed by prior works, and improving the complexity of $k$-class multi-calibration by an exponential factor of $k$ versus Gopalan et al.. Beyond multi-calibration, we use these game dynamics to address emerging considerations in the study of group fairness and multi-distribution learning.

We consider the problem of best arm identification in the multi-armed bandit model, under fixed confidence. Given a confidence input $\delta$, the goal is to identify the arm with the highest mean reward with a probability of at least $1 - \delta$, while minimizing the number of arm pulls. While the literature provides solutions to this problem under the assumption of independent arms distributions, we propose a more flexible scenario where arms can be dependent and rewards can be sampled simultaneously. This framework allows the learner to estimate the covariance among the arms distributions, enabling a more efficient identification of the best arm. The relaxed setting we propose is relevant in various applications, such as clinical trials, where similarities between patients or drugs suggest underlying correlations in the outcomes. We introduce new algorithms that adapt to the unknown covariance of the arms and demonstrate through theoretical guarantees that substantial improvement can be achieved over the standard setting. Additionally, we provide new lower bounds for the relaxed setting and present numerical simulations that support their theoretical findings.

It is a crucial challenge to reconstruct population dynamics using unlabeled samples from distributions at coarse time intervals. Recent approaches such as flow-based models or Schrödinger Bridge (SB) models have demonstrated appealing performance, yet the inferred sample trajectories either fail to account for the underlying stochasticity or are unnecessarily rigid. In this article, we extend SB into phase space and propose $\underline{D}$eep $\underline{M}$omentum Multi-Marginal $\underline{S}$chrödinger $\underline{B}$ridge (DMSB), a novel computational framework that learns the smooth measure-valued spline for stochastic systems that satisfy position marginal constraints across time. By tailoring the celebrated Bregman Iteration and extending the Iteration Proportional Fitting to phase space, we manage to handle high-dimensional multi-marginal trajectory inference tasks efficiently. Our algorithm outperforms baselines significantly, as evidenced by experiments for synthetic datasets and a real-world single-cell RNA sequence dataset. Additionally, the proposed approach can reasonably reconstruct the evolution of velocity distribution, from position snapshots only, when there is a ground truth velocity that is nevertheless inaccessible.

Whittle index policy is a heuristic to the intractable restless multi-armed bandits (RMAB) problem. Although it is provably asymptotically optimal, finding Whittle indices remains difficult. In this paper, we present Neural-Q-Whittle, a Whittle index based Q-learning algorithm for RMAB with neural network function approximation, which is an example of nonlinear two-timescale stochastic approximation with Q-function values updated on a faster timescale and Whittle indices on a slower timescale. Despite the empirical success of deep Q-learning, the non-asymptotic convergence rate of Neural-Q-Whittle, which couples neural networks with two-timescale Q-learning largely remains unclear. This paper provides a finite-time analysis of Neural-Q-Whittle, where data are generated from a Markov chain, and Q-function is approximated by a ReLU neural network. Our analysis leverages a Lyapunov drift approach to capture the evolution of two coupled parameters, and the nonlinearity in value function approximation further requires us to characterize the approximation error. Combing these provide Neural-Q-Whittle with $\mathcal{O}(1/k^{2/3})$ convergence rate, where $k$ is the number of iterations.

In an era of more diverse data modalities, multi-view clustering has become a fundamental tool for comprehensive data analysis and exploration. However, existing multi-view unsupervised learning methods often rely on strict assumptions on semantic consistency among samples. In this paper, we reformulate the multi-view clustering problem from an information-theoretic perspective and propose a general theoretical model. In particular, we define three desiderata under multi-view unsupervised learning in terms of mutual information, namely, comprehensiveness, concentration, and cross-diversity. The multi-view variational lower bound is then obtained by approximating the samples' high-dimensional mutual information. The Kullback–Leibler divergence is utilized to deduce sample assignments. Ultimately the information-based multi-view clustering model leverages deep neural networks and Stochastic Gradient Variational Bayes to achieve representation learning and clustering simultaneously. Extensive experiments on both synthetic and real datasets with wide types demonstrate that the proposed method exhibits a more stable and superior clustering performance than state-of-the-art algorithms.

Skill discovery has shown significant progress in unsupervised reinforcement learning. This approach enables the discovery of a wide range of skills without any extrinsic reward, which can be effectively combined to tackle complex tasks. However, such unsupervised skill learning has not been well applied to multi-agent reinforcement learning (MARL) due to two primary challenges. One is how to learn skills not only for the individual agents but also for the entire team, and the other is how to coordinate the skills of different agents to accomplish multi-agent tasks. To address these challenges, we present Hierarchical Multi-Agent Skill Discovery (HMASD), a two-level hierarchical algorithm for discovering both team and individual skills in MARL. The high-level policy employs a transformer structure to realize sequential skill assignment, while the low-level policy learns to discover valuable team and individual skills. We evaluate HMASD on sparse reward multi-agent benchmarks, and the results show that HMASD achieves significant performance improvements compared to strong MARL baselines.

We study $K$-armed bandit problems where the reward distributions of the arms are all supported on the $[0,1]$ interval. Maillard sampling\cite{maillard13apprentissage}, an attractive alternative to Thompson sampling, has recently been shown to achieve competitive regret guarantees in the sub-Gaussian reward setting\cite{bian2022maillard} while maintaining closed-form action probabilities, which is useful for offline policy evaluation. In this work, we analyze the Kullback-Leibler Maillard Sampling (KL-MS) algorithm, a natural extension of Maillard sampling {and a special case of Minimum Empirical Divergence (MED)~\cite{honda2011asymptotically}} for achieving a KL-style finite-time gap-dependent regret bound. We show that KL-MS enjoys the asymptotic optimality when the rewards are Bernoulli and has an {adaptive} worst-case regret bound of the form $O(\sqrt{\mu^*(1-\mu^*) K T \ln K} + K \ln T)$, where $\mu^*$ is the expected reward of the optimal arm, and $T$ is the time horizon length; {this is the first time such adaptivity is reported in the literature for an algorithm with asymptotic optimality guarantees.}

We study the problem of designing adaptive multi-armed bandit algorithms that perform optimally in both the stochastic setting and the adversarial setting simultaneously (often known as a best-of-both-world guarantee). A line of recent works shows that when configured and analyzed properly, the Follow-the-Regularized-Leader (FTRL) algorithm, originally designed for the adversarial setting, can in fact optimally adapt to the stochastic setting as well. Such results, however, critically rely on an assumption that there exists one unique optimal arm. Recently, Ito [2021] took the first step to remove such an undesirable uniqueness assumption for one particular FTRL algorithm with the 1/2-Tsallis entropy regularizer. In this work, we significantly improve and generalize this result, showing that uniqueness is unnecessary for FTRL with a broad family of regularizers and a new learning rate schedule. For some regularizers, our regret bounds also improve upon prior results even when uniqueness holds. We further provide an application of our results to the decoupled exploration and exploitation problem, demonstrating that our techniques are broadly applicable.

The framework of multi-agent learning explores the dynamics of how an agent's strategies evolve in response to the evolving strategies of other agents. Of particular interest is whether or not agent strategies converge to well known solution concepts such as Nash Equilibrium (NE). In "higher order'' learning, agent dynamics include auxiliary states that can capture phenomena such as path dependencies. We introduce higher-order gradient play dynamics that resemble projected gradient ascent with auxiliary states. The dynamics are "payoff based'' and "uncoupled'' in that each agent's dynamics depend on its own evolving payoff and has no explicit dependence on the utilities of other agents. We first show that for any specific game with an isolated completely mixed-strategy NE, there exist higher-order gradient play dynamics that lead (locally) to that NE, both for the specific game and nearby games with perturbed utility functions. Conversely, we show that for any higher-order gradient play dynamics, there exists a game with a unique isolated completely mixed-strategy NE for which the dynamics do not lead to NE. Finally, we show that convergence to the mixed-strategy equilibrium in coordination games, comes at the expense of the dynamics being inherently internally unstable.

Multi-view clustering (MVC) is a popular technique for improving clustering performance using various data sources. However, existing methods primarily focus on acquiring consistent information while often neglecting the issue of redundancy across multiple views. This study presents a new approach called Sufficient Multi-View Clustering (SUMVC) that examines the multi-view clustering framework from an information-theoretic standpoint. Our proposed method consists of two parts. Firstly, we develop a simple and reliable multi-view clustering method SCMVC (simple consistent multi-view clustering) that employs variational analysis to generate consistent information. Secondly, we propose a sufficient representation lower bound to enhance consistent information and minimise unnecessary information among views. The proposed SUMVC method offers a promising solution to the problem of multi-view clustering and provides a new perspective for analyzing multi-view data. To verify the effectiveness of our model, we conducted a theoretical analysis based on the Bayes Error Rate, and experiments on multiple multi-view datasets demonstrate the superior performance of SUMVC.

The lack of standardized benchmarks for reinforcement learning (RL) in sustainability applications has made it difficult to both track progress on specific domains and identify bottlenecks for researchers to focus their efforts. In this paper, we present SustainGym, a suite of five environments designed to test the performance of RL algorithms on realistic sustainable energy system tasks, ranging from electric vehicle charging to carbon-aware data center job scheduling. The environments test RL algorithms under realistic distribution shifts as well as in multi-agent settings. We show that standard off-the-shelf RL algorithms leave significant room for improving performance and highlight the challenges ahead for introducing RL to real-world sustainability tasks.

Recognizing and organizing text in correct reading order plays a crucial role in historical document analysis and preservation. While existing methods have shown promising performance, they often struggle with challenges such as diverse layouts, low image quality, style variations, and distortions. This is primarily due to the lack of consideration for these issues in the current benchmarks, which hinders the development and evaluation of historical document analysis and recognition (HDAR) methods in complex real-world scenarios. To address this gap, this paper introduces a complex multi-style Chinese historical document analysis benchmark, named M5HisDoc. The M5 indicates five properties of style, ie., Multiple layouts, Multiple document types, Multiple calligraphy styles, Multiple backgrounds, and Multiple challenges. The M5HisDoc dataset consists of two subsets, M5HisDoc-R (Regular) and M5HisDoc-H (Hard). The M5HisDoc-R subset comprises 4,000 historical document images. To ensure high-quality annotations, we meticulously perform manual annotation and triple-checking. To replicate real-world conditions for historical document analysis applications, we incorporate image rotation, distortion, and resolution reduction into M5HisDoc-R subset to form a new challenging subset named M5HisDoc-H, which contains the same number of images as M5HisDoc-R. The dataset exhibits diverse styles, significant scale variations, dense texts, and an extensive character set. We conduct benchmarking experiments on five tasks: text line detection, text line recognition, character detection, character recognition, and reading order prediction. We also conduct cross-validation with other benchmarks. Experimental results demonstrate that the M5HisDoc dataset can offer new challenges and great opportunities for future research in this field, thereby providing deep insights into the solution for HDAR. The dataset is available at https://github.com/HCIILAB/M5HisDoc.

Modern recording techniques now allow us to record from distinct neuronal populations in different brain networks. However, especially as we consider multiple (more than two) populations, new conceptual and statistical frameworks are needed to characterize the multi-dimensional, concurrent flow of signals among these populations. Here, we develop a dimensionality reduction framework that determines (1) the subset of populations described by each latent dimension, (2) the direction of signal flow among those populations, and (3) how those signals evolve over time within and across experimental trials. We illustrate these features in simulation, and further validate the method by applying it to previously studied recordings from neuronal populations in macaque visual areas V1 and V2. Then we study interactions across select laminar compartments of areas V1, V2, and V3d, recorded simultaneously with multiple Neuropixels probes. Our approach uncovered signatures of selective communication across these three areas that related to their retinotopic alignment. This work advances the study of concurrent signaling across multiple neuronal populations.

Neural policy learning methods have achieved remarkable results in various control problems, ranging from Atari games to simulated locomotion. However, these methods struggle in long-horizon tasks, especially in open-ended environments with multi-modal observations, such as the popular dungeon-crawler game, NetHack. Intriguingly, the NeurIPS 2021 NetHack Challenge revealed that symbolic agents outperformed neural approaches by over four times in median game score. In this paper, we delve into the reasons behind this performance gap and present an extensive study on neural policy learning for NetHack. To conduct this study, we analyze the winning symbolic agent, extending its codebase to track internal strategy selection in order to generate one of the largest available demonstration datasets. Utilizing this dataset, we examine (i) the advantages of an action hierarchy; (ii) enhancements in neural architecture; and (iii) the integration of reinforcement learning with imitation learning. Our investigations produce a state-of-the-art neural agent that surpasses previous fully neural policies by 127% in offline settings and 25% in online settings on median game score. However, we also demonstrate that mere scaling is insufficient to bridge the performance gap with the best symbolic models or even the top human players.

Fictitious Play (FP) is a simple and natural dynamic for repeated play with many applications in game theory and multi-agent reinforcement learning. It was introduced by Brown and its convergence properties for two-player zero-sum games was established later by Robinson. Potential games [Monderer and Shapley 1996] is another class of games which exhibit the FP property [Monderer and Shapley 1996], i.e., FP dynamics converges to a Nash equilibrium if all agents follows it. Nevertheless, except for two-player zero-sum games and for specific instances of payoff matrices [Abernethy et. al. 2021] or for adversarial tie-breaking rules [Daskalakis and Pan, 2014], the \textit{convergence rate} of FP is unknown. In this work, we focus on the rate of convergence of FP when applied to potential games and more specifically identical payoff games. We prove that FP can take exponential time (in the number of strategies) to reach a Nash equilibrium, even if the game is restricted to \textit{two agents}. To prove this, we recursively construct a two-player coordination game with a unique Nash equilibrium. Moreover, every approximate Nash equilibrium in the constructed game must be close to the pure Nash equilibrium in $\ell_1$-distance.

High sample complexity has long been a challenge for RL. On the other hand, humans learn to perform tasks not only from interaction or demonstrations, but also by reading unstructured text documents, e.g., instruction manuals. Instruction manuals and wiki pages are among the most abundant data that could inform agents of valuable features and policies or task-specific environmental dynamics and reward structures. Therefore, we hypothesize that the ability to utilize human-written instruction manuals to assist learning policies for specific tasks should lead to a more efficient and better-performing agent. We propose the Read and Reward framework. Read and Reward speeds up RL algorithms on Atari games by reading manuals released by the Atari game developers. Our framework consists of a QA Extraction module that extracts and summarizes relevant information from the manual and a Reasoning module that evaluates object-agent interactions based on information from the manual. An auxiliary reward is then provided to a standard A2C RL agent, when interaction is detected. Experimentally, various RL algorithms obtain significant improvement in performance and training speed when assisted by our design. Code at github.com/Holmeswww/RnR

We introduce IMP-MARL, an open-source suite of multi-agent reinforcement learning (MARL) environments for large-scale Infrastructure Management Planning (IMP), offering a platform for benchmarking the scalability of cooperative MARL methods in real-world engineering applications. In IMP, a multi-component engineering system is subject to a risk of failure due to its components' damage condition. Specifically, each agent plans inspections and repairs for a specific system component, aiming to minimise maintenance costs while cooperating to minimise system failure risk. With IMP-MARL, we release several environments including one related to offshore wind structural systems, in an effort to meet today's needs to improve management strategies to support sustainable and reliable energy systems. Supported by IMP practical engineering environments featuring up to 100 agents, we conduct a benchmark campaign, where the scalability and performance of state-of-the-art cooperative MARL methods are compared against expert-based heuristic policies. The results reveal that centralised training with decentralised execution methods scale better with the number of agents than fully centralised or decentralised RL approaches, while also outperforming expert-based heuristic policies in most IMP environments. Based on our findings, we additionally outline remaining cooperation and scalability challenges that future MARL methods should still address. Through IMP-MARL, we encourage the implementation of new environments and the further development of MARL methods.