As a primary contribution, we present a convergence theorem for stochastic iterations, and in particular, Q-learning iterates, under a general, possibly non-Markovian, stochastic environment. Our conditions for convergence involve an ergodicity and a positivity criterion. We provide a precise characterization on the limit of the iterates and conditions on the environment and initializations for convergence. As our second contribution, we discuss the implications and applications of this theorem to a variety of stochastic control problems with non-Markovian environments involving (i) quantized approximations of fully observed Markov Decision Processes (MDPs) with continuous spaces (where quantization break down the Markovian structure), (ii) quantized approximations of belief-MDP reduced partially observable MDPS (POMDPs) with weak Feller continuity and a mild version of filter stability (which requires the knowledge of the model by the controller), (iii) finite window approximations of POMDPs under a uniform controlled filter stability (which does not require the knowledge of the model), and (iv) for multi-agent models where convergence of learning dynamics to a new class of equilibria, subjective Q-learning equilibria, will be studied. In addition to the convergence theorem, some implications of the theorem above are new to the literature and others are interpreted as applications of the convergence theorem. Some open problems are noted.

Offline reinforcement learning (RL) can in principle synthesize more optimal behavior from a dataset consisting only of suboptimal trials. One way that this can happen is by "stitching" together the best parts of otherwise suboptimal trajectories that overlap on similar states, to create new behaviors where each individual state is in-distribution, but the overall returns are higher. However, in many interesting and complex applications, such as autonomous navigation and dialogue systems, the state is partially observed. Even worse, the state representation is unknown or not easy to define. In such cases, policies and value functions are often conditioned on observation histories instead of states. In these cases, it is not clear if the same kind of "stitching" is feasible at the level of observation histories, since two different trajectories would always have different histories, and thus "similar states" that might lead to effective stitching cannot be leveraged. Theoretically, we show that standard offline RL algorithms conditioned on observation histories suffer from poor sample complexity, in accordance with the above intuition. We then identify sufficient conditions under which offline RL can still be efficient -- intuitively, it needs to learn a compact representation of history comprising only features relevant for action selection. We introduce a bisimulation loss that captures the extent to which this happens, and propose that offline RL can explicitly optimize this loss to aid worst-case sample complexity. Empirically, we show that across a variety of tasks either our proposed loss improves performance, or the value of this loss is already minimized as a consequence of standard offline RL, indicating that it correlates well with good performance.

Robust multi-agent reinforcement learning (MARL) necessitates resilience to uncertain or worst-case actions by unknown allies. Existing max-min optimization techniques in robust MARL seek to enhance resilience by training agents against worst-case adversaries, but this becomes intractable as the number of agents grows, leading to exponentially increasing worst-case scenarios. Attempts to simplify this complexity often yield overly pessimistic policies, inadequate robustness across scenarios and high computational demands. Unlike these approaches, humans naturally learn adaptive and resilient behaviors without the necessity of preparing for every conceivable worst-case scenario. Motivated by this, we propose MIR2, which trains policy in routine scenarios and minimize Mutual Information as Robust Regularization. Theoretically, we frame robustness as an inference problem and prove that minimizing mutual information between histories and actions implicitly maximizes a lower bound on robustness under certain assumptions. Further analysis reveals that our proposed approach prevents agents from overreacting to others through an information bottleneck and aligns the policy with a robust action prior. Empirically, our MIR2 displays even greater resilience against worst-case adversaries than max-min optimization in StarCraft II, Multi-agent Mujoco and rendezvous. Our superiority is consistent when deployed in challenging real-world robot swarm control scenario. See code and demo videos in Supplementary Materials.

NoMoPy is a code for fitting, analyzing, and generating noise modeled as a hidden Markov model (HMM) or, more generally, factorial hidden Markov model (FHMM). This code, written in Python, implements approximate and exact expectation maximization (EM) algorithms for performing the parameter estimation process, model selection procedures via cross-validation, and parameter confidence region estimation. Here, we describe in detail the functionality implemented in NoMoPy and provide examples of its use and performance on example problems.

To make reinforcement learning more sample efficient, we need better credit assignment methods that measure an action's influence on future rewards. Building upon Hindsight Credit Assignment (HCA), we introduce Counterfactual Contribution Analysis (COCOA), a new family of model-based credit assignment algorithms. Our algorithms achieve precise credit assignment by measuring the contribution of actions upon obtaining subsequent rewards, by quantifying a counterfactual query: 'Would the agent still have reached this reward if it had taken another action?'. We show that measuring contributions w.r.t. rewarding states, as is done in HCA, results in spurious estimates of contributions, causing HCA to degrade towards the high-variance REINFORCE estimator in many relevant environments. Instead, we measure contributions w.r.t. rewards or learned representations of the rewarding objects, resulting in gradient estimates with lower variance. We run experiments on a suite of problems specifically designed to evaluate long-term credit assignment capabilities. By using dynamic programming, we measure ground-truth policy gradients and show that the improved performance of our new model-based credit assignment methods is due to lower bias and variance compared to HCA and common baselines. Our results demonstrate how modeling action contributions towards rewarding outcomes can be leveraged for credit assignment, opening a new path towards sample-efficient reinforcement learning.

Visual reinforcement learning (RL) has shown promise in continuous control tasks. Despite its progress, current algorithms are still unsatisfactory in virtually every aspect of the performance such as sample efficiency, asymptotic performance, and their robustness to the choice of random seeds. In this paper, we identify a major shortcoming in existing visual RL methods that is the agents often exhibit sustained inactivity during early training, thereby limiting their ability to explore effectively. Expanding upon this crucial observation, we additionally unveil a significant correlation between the agents' inclination towards motorically inactive exploration and the absence of neuronal activity within their policy networks. To quantify this inactivity, we adopt dormant ratio as a metric to measure inactivity in the RL agent's network. Empirically, we also recognize that the dormant ratio can act as a standalone indicator of an agent's activity level, regardless of the received reward signals. Leveraging the aforementioned insights, we introduce DrM, a method that uses three core mechanisms to guide agents' exploration-exploitation trade-offs by actively minimizing the dormant ratio. Experiments demonstrate that DrM achieves significant improvements in sample efficiency and asymptotic performance with no broken seeds (76 seeds in total) across three continuous control benchmark environments, including DeepMind Control Suite, MetaWorld, and Adroit. Most importantly, DrM is the first model-free algorithm that consistently solves tasks in both the Dog and Manipulator domains from the DeepMind Control Suite as well as three dexterous hand manipulation tasks without demonstrations in Adroit, all based on pixel observations.

Humans write code in a fundamentally interactive manner and rely on constant execution feedback to correct errors, resolve ambiguities, and decompose tasks. While LLMs have recently exhibited promising coding capabilities, current coding benchmarks mostly consider a static instruction-to-code sequence transduction process, which has the potential for error propagation and a disconnect between the generated code and its final execution environment. To address this gap, we introduce InterCode, a lightweight, flexible, and easy-to-use framework of interactive coding as a standard reinforcement learning (RL) environment, with code as actions and execution feedback as observations. Our framework is language and platform agnostic, uses self-contained Docker environments to provide safe and reproducible execution, and is compatible out-of-the-box with traditional seq2seq coding methods, while enabling the development of new methods for interactive code generation. We use InterCode to create three interactive code environments with Bash, SQL, and Python as action spaces, leveraging data from the static NL2Bash [32], Spider [55], and MBPP [4] datasets. We demonstrate InterCode's viability as a testbed by evaluating multiple state-of-the-art LLMs configured with different prompting strategies such as ReAct [51] and Plan & Solve [43]. Our results showcase the benefits of interactive code generation and demonstrate that InterCode can serve as a challenging benchmark for advancing code understanding and generation capabilities. InterCode is designed to be easily extensible and can even be used to create new tasks such as Capture the Flag, a popular coding puzzle that is inherently multi-step and involves multiple programming languages.

Optimal plans in Constrained Partially Observable Markov Decision Processes (CPOMDPs) maximize reward objectives while satisfying hard cost constraints, generalizing safe planning under state and transition uncertainty. Unfortunately, online CPOMDP planning is extremely difficult in large or continuous problem domains. In many large robotic domains, hierarchical decomposition can simplify planning by using tools for low-level control given high-level action primitives (options). We introduce Constrained Options Belief Tree Search (COBeTS) to leverage this hierarchy and scale online search-based CPOMDP planning to large robotic problems. We show that if primitive option controllers are defined to satisfy assigned constraint budgets, then COBeTS will satisfy constraints anytime. Otherwise, COBeTS will guide the search towards a safe sequence of option primitives, and hierarchical monitoring can be used to achieve runtime safety. We demonstrate COBeTS in several safety-critical, constrained partially observable robotic domains, showing that it can plan successfully in continuous CPOMDPs while non-hierarchical baselines cannot.

Computational models and simulations are not just appealing because of their intrinsic characteristics across spatiotemporal scales, scalability, and predictive power, but also because the set of problems in cancer biomedicine that can be addressed computationally exceeds the set of those amenable to analytical solutions. Agent-based models and simulations are especially interesting candidates among computational modelling strategies in cancer research due to their capabilities to replicate realistic local and global interaction dynamics at a convenient and relevant scale. Yet, the absence of methods to validate the consistency of the results across scales can hinder adoption by turning fine-tuned models into black boxes. This review compiles relevant literature to explore strategies to leverage high-fidelity simulations of multi-scale, or multi-level, cancer models with a focus on validation approached as simulation calibration. We argue that simulation calibration goes beyond parameter optimization by embedding informative priors to generate plausible parameter configurations across multiple dimensions.

Interaction-aware Autonomous Driving (IAAD) is a rapidly growing field of research that focuses on the development of autonomous vehicles (AVs) that are capable of interacting safely and efficiently with human road users. This is a challenging task, as it requires the autonomous vehicle to be able to understand and predict the behaviour of human road users. In this literature review, the current state of IAAD research is surveyed in this work. Commencing with an examination of terminology, attention is drawn to challenges and existing models employed for modelling the behaviour of drivers and pedestrians. Next, a comprehensive review is conducted on various techniques proposed for interaction modelling, encompassing cognitive methods, machine learning approaches, and game-theoretic methods. The conclusion is reached through a discussion of potential advantages and risks associated with IAAD, along with the illumination of pivotal research inquiries necessitating future exploration.