Automatic bill payment is an important part of business operations in fintech companies. The practice of deduction was mainly based on the total amount or heuristic search by dividing the bill into smaller parts to deduct as much as possible. This article proposes an end-to-end approach of automatically learning the optimal deduction paths (deduction amount in order), which reduces the cost of manual path design and maximizes the amount of successful deduction. Specifically, in view of the large search space of the paths and the extreme sparsity of historical successful deduction records, we propose a deep hierarchical reinforcement learning approach which abstracts the action into a two-level hierarchical space: an upper agent that determines the number of steps of deductions each day and a lower agent that decides the amount of deduction at each step. In such a way, the action space is structured via prior knowledge and the exploration space is reduced. Moreover, the inherited information incompleteness of the business makes the environment just partially observable. To be precise, the deducted amounts indicate merely the lower bounds of the available account balance. To this end, we formulate the problem as a partially observable Markov decision problem (POMDP) and employ an environment correction algorithm based on the characteristics of the business. In the world's largest electronic payment business, we have verified the effectiveness of this scheme offline and deployed it online to serve millions of users.

In reinforcement learning (RL), there are two major settings for interacting with the environment: online and offline. Online methods explore the environment at significant time cost, and offline methods efficiently obtain reward signals by sacrificing exploration capability. We propose semi-offline RL, a novel paradigm that smoothly transits from offline to online settings, balances exploration capability and training cost, and provides a theoretical foundation for comparing different RL settings. Based on the semi-offline formulation, we present the RL setting that is optimal in terms of optimization cost, asymptotic error, and overfitting error bound. Extensive experiments show that our semi-offline approach is efficient and yields comparable or often better performance compared with state-of-the-art methods.

Learning interpretable representations of neural dynamics at a population level is a crucial first step to understanding how observed neural activity relates to perception and behavior. Models of neural dynamics often focus on either low-dimensional projections of neural activity, or on learning dynamical systems that explicitly relate to the neural state over time. We discuss how these two approaches are interrelated by considering dynamical systems as representative of flows on a low-dimensional manifold. Building on this concept, we propose a new decomposed dynamical system model that represents complex non-stationary and nonlinear dynamics of time series data as a sparse combination of simpler, more interpretable components. Our model is trained through a dictionary learning procedure, where we leverage recent results in tracking sparse vectors over time. The decomposed nature of the dynamics is more expressive than previous switched approaches for a given number of parameters and enables modeling of overlapping and non-stationary dynamics. In both continuous-time and discrete-time instructional examples we demonstrate that our model can well approximate the original system, learn efficient representations, and capture smooth transitions between dynamical modes, focusing on intuitive low-dimensional non-stationary linear and nonlinear systems. Furthermore, we highlight our model's ability to efficiently capture and demix population dynamics generated from multiple independent subnetworks, a task that is computationally impractical for switched models. Finally, we apply our model to neural "full brain" recordings of C. elegans data, illustrating a diversity of dynamics that is obscured when classified into discrete states.

We address the problem of monitoring a set of binary stochastic processes and generating an alert when the number of anomalies among them exceeds a threshold. For this, the decision-maker selects and probes a subset of the processes to obtain noisy estimates of their states (normal or anomalous). Based on the received observations, the decisionmaker first determines whether to declare that the number of anomalies has exceeded the threshold or to continue taking observations. When the decision is to continue, it then decides whether to collect observations at the next time instant or defer it to a later time. If it chooses to collect observations, it further determines the subset of processes to be probed. To devise this three-step sequential decision-making process, we use a Bayesian formulation wherein we learn the posterior probability on the states of the processes. Using the posterior probability, we construct a Markov decision process and solve it using deep actor-critic reinforcement learning. Via numerical experiments, we demonstrate the superior performance of our algorithm compared to the traditional model-based algorithms.

In-context learning (ICL) is one of the most powerful and most unexpected capabilities to emerge in recent transformer-based large language models (LLMs). Yet the mechanisms that underlie it are poorly understood. In this paper, we demonstrate that comparable ICL capabilities can be acquired by an alternative sequence prediction learning method using clone-structured causal graphs (CSCGs). Moreover, a key property of CSCGs is that, unlike transformer-based LLMs, they are {\em interpretable}, which considerably simplifies the task of explaining how ICL works. Specifically, we show that it uses a combination of (a) learning template (schema) circuits for pattern completion, (b) retrieving relevant templates in a context-sensitive manner, and (c) rebinding of novel tokens to appropriate slots in the templates. We go on to marshall evidence for the hypothesis that similar mechanisms underlie ICL in LLMs. For example, we find that, with CSCGs as with LLMs, different capabilities emerge at different levels of overparameterization, suggesting that overparameterization helps in learning more complex template (schema) circuits. By showing how ICL can be achieved with small models and datasets, we open up a path to novel architectures, and take a vital step towards a more general understanding of the mechanics behind this important capability.

Multi-task missions for unmanned aerial vehicles (UAVs) involving inspection and landing tasks are challenging for novice pilots due to the difficulties associated with depth perception and the control interface. We propose a shared autonomy system, alongside supplementary information displays, to assist pilots to successfully complete multi-task missions without any pilot training. Our approach comprises of three modules: (1) a perception module that encodes visual information onto a latent representation, (2) a policy module that augments pilot's actions, and (3) an information augmentation module that provides additional information to the pilot. The policy module is trained in simulation with simulated users and transferred to the real world without modification in a user study (n=29), alongside supplementary information schemes including learnt red/green light feedback cues and an augmented reality display. The pilot's intent is unknown to the policy module and is inferred from the pilot's input and UAV's states. The assistant increased task success rate for the landing and inspection tasks from [16.67% & 54.29%] respectively to [95.59% & 96.22%]. With the assistant, inexperienced pilots achieved similar performance to experienced pilots. Red/green light feedback cues reduced the required time by 19.53% and trajectory length by 17.86% for the inspection task, where participants rated it as their preferred condition due to the intuitive interface and providing reassurance. This work demonstrates that simple user models can train shared autonomy systems in simulation, and transfer to physical tasks to estimate user intent and provide effective assistance and information to the pilot.

There has been a recent surge of interest in developing generally-capable agents that can adapt to new tasks without additional training in the environment. Learning world models from reward-free exploration is a promising approach, and enables policies to be trained using imagined experience for new tasks. Achieving a general agent requires robustness across different environments. However, different environments may require different amounts of data to learn a suitable world model. In this work, we address the problem of efficiently learning robust world models in the reward-free setting. As a measure of robustness, we consider the minimax regret objective. We show that the minimax regret objective can be connected to minimising the maximum error in the world model across environments. This informs our algorithm, WAKER: Weighted Acquisition of Knowledge across Environments for Robustness. WAKER selects environments for data collection based on the estimated error of the world model for each environment. Our experiments demonstrate that WAKER outperforms naive domain randomisation, resulting in improved robustness, efficiency, and generalisation.

Although powerful graph neural networks (GNNs) have boosted numerous real-world applications, the potential privacy risk is still underexplored. To close this gap, we perform the first comprehensive study of graph reconstruction attack that aims to reconstruct the adjacency of nodes. We show that a range of factors in GNNs can lead to the surprising leakage of private links. Especially by taking GNNs as a Markov chain and attacking GNNs via a flexible chain approximation, we systematically explore the underneath principles of graph reconstruction attack, and propose two information theory-guided mechanisms: (1) the chain-based attack method with adaptive designs for extracting more private information; (2) the chain-based defense method that sharply reduces the attack fidelity with moderate accuracy loss. Such two objectives disclose a critical belief that to recover better in attack, you must extract more multi-aspect knowledge from the trained GNN; while to learn safer for defense, you must forget more link-sensitive information in training GNNs. Empirically, we achieve state-of-the-art results on six datasets and three common GNNs. The code is publicly available at: https://github.com/tmlr-group/MC-GRA.

Offline reinforcement learning (RL) that learns policies from offline datasets without environment interaction has received considerable attention in recent years. Compared with the rich literature in the single-agent case, offline multi-agent RL is still a relatively underexplored area. Most existing methods directly apply offline RL ingredients in the multi-agent setting without fully leveraging the decomposable problem structure, leading to less satisfactory performance in complex tasks. We present OMAC, a new offline multi-agent RL algorithm with coupled value factorization. OMAC adopts a coupled value factorization scheme that decomposes the global value function into local and shared components, and also maintains the credit assignment consistency between the state-value and Q-value functions. Moreover, OMAC performs in-sample learning on the decomposed local state-value functions, which implicitly conducts max-Q operation at the local level while avoiding distributional shift caused by evaluating out-of-distribution actions. Based on the comprehensive evaluations of the offline multi-agent StarCraft II micro-management tasks, we demonstrate the superior performance of OMAC over the state-of-the-art offline multi-agent RL methods.

Decentralized multi-agent reinforcement learning (MARL) algorithms have become popular in the literature since it allows heterogeneous agents to have their own reward functions as opposed to canonical multi-agent Markov Decision Process (MDP) settings which assume common reward functions over all agents. In this work, we follow the existing work on collaborative MARL where agents in a connected time varying network can exchange information among each other in order to reach a consensus. We introduce vulnerabilities in the consensus updates of existing MARL algorithms where agents can deviate from their usual consensus update, who we term as adversarial agents. We then proceed to provide an algorithm that allows non-adversarial agents to reach a consensus in the presence of adversaries under a constrained setting.