Offline meta-reinforcement learning aims to equip agents with the ability to rapidly adapt to new tasks by training on data from a set of different tasks. Context-based approaches utilize a history of state-action-reward transitions -- referred to as the context -- to infer representations of the current task, and then condition the agent, i.e., the policy and value function, on the task representations. Intuitively, the better the task representations capture the underlying tasks, the better the agent can generalize to new tasks. Unfortunately, context-based approaches suffer from distribution mismatch, as the context in the offline data does not match the context at test time, limiting their ability to generalize to the test tasks. This leads to the task representations overfitting to the offline training data. Intuitively, the task representations should be independent of the behavior policy used to collect the offline data. To address this issue, we approximately minimize the mutual information between the distribution over the task representations and behavior policy by maximizing the entropy of behavior policy conditioned on the task representations. We validate our approach in MuJoCo environments, showing that compared to baselines, our task representations more faithfully represent the underlying tasks, leading to outperforming prior methods in both in-distribution and out-of-distribution tasks.
Meta-Reinforcement Learning (Meta-RL) enables fast adaptation to new testing tasks. Despite recent advancements, it is still challenging to learn performant policies across multiple complex and high-dimensional tasks. To address this, we propose a novel architecture with three hierarchical levels for 1) learning task representations, 2) discovering task-agnostic macro-actions in an automated manner, and 3) learning primitive actions. The macro-action can guide the low-level primitive policy learning to more efficiently transition to goal states. This can address the issue that the policy may forget previously learned behavior while learning new, conflicting tasks. Moreover, the task-agnostic nature of the macro-actions is enabled by removing task-specific components from the state space. Hence, this makes them amenable to re-composition across different tasks and leads to promising fast adaptation to new tasks. Also, the prospective instability from the tri-level hierarchies is effectively mitigated by our innovative, independently tailored training schemes. Experiments in the MetaWorld framework demonstrate the improved sample efficiency and success rate of our approach compared to previous state-of-the-art methods.
With increasing numbers of vulnerabilities exposed on the internet, autonomous penetration testing (pentesting) has emerged as a promising research area. Reinforcement learning (RL) is a natural fit for studying this topic. However, two key challenges limit the applicability of RL-based autonomous pentesting in real-world scenarios: (a) training environment dilemma -- training agents in simulated environments is sample-efficient while ensuring their realism remains challenging; (b) poor generalization ability -- agents' policies often perform poorly when transferred to unseen scenarios, with even slight changes potentially causing significant generalization gap. To this end, we propose GAP, a generalizable autonomous pentesting framework that aims to realizes efficient policy training in realistic environments and train generalizable agents capable of drawing inferences about other cases from one instance. GAP introduces a Real-to-Sim-to-Real pipeline that (a) enables end-to-end policy learning in unknown real environments while constructing realistic simulations; (b) improves agents' generalization ability by leveraging domain randomization and meta-RL learning.Specially, we are among the first to apply domain randomization in autonomous pentesting and propose a large language model-powered domain randomization method for synthetic environment generation. We further apply meta-RL to improve agents' generalization ability in unseen environments by leveraging synthetic environments. The combination of two methods effectively bridges the generalization gap and improves agents' policy adaptation performance.Experiments are conducted on various vulnerable virtual machines, with results showing that GAP can enable policy learning in various realistic environments, achieve zero-shot policy transfer in similar environments, and realize rapid policy adaptation in dissimilar environments.
We discuss the possibility of world models and active exploration as emergent properties of open-ended behavior optimization in autonomous agents. In discussing the source of the open-endedness of living things, we start from the perspective of biological systems as understood by the mechanistic approach of theoretical biology and artificial life. From this perspective, we discuss the potential of homeostasis in particular as an open-ended objective for autonomous agents and as a general, integrative extrinsic motivation. We then discuss the possibility of implicitly acquiring a world model and active exploration through the internal dynamics of a network, and a hypothetical architecture for this, by combining meta-reinforcement learning, which assumes domain adaptation as a system that achieves robust homeostasis.
Language models trained on diverse datasets unlock generalization by in-context learning. Reinforcement Learning (RL) policies can achieve a similar effect by meta-learning within the memory of a sequence model. However, meta-RL research primarily focuses on adapting to minor variations of a single task. It is difficult to scale towards more general behavior without confronting challenges in multi-task optimization, and few solutions are compatible with meta-RL's goal of learning from large training sets of unlabeled tasks. To address this challenge, we revisit the idea that multi-task RL is bottlenecked by imbalanced training losses created by uneven return scales across different tasks. We build upon recent advancements in Transformer-based (in-context) meta-RL and evaluate a simple yet scalable solution where both an agent's actor and critic objectives are converted to classification terms that decouple optimization from the current scale of returns. Large-scale comparisons in Meta-World ML45, Multi-Game Procgen, Multi-Task POPGym, Multi-Game Atari, and BabyAI find that this design unlocks significant progress in online multi-task adaptation and memory problems without explicit task labels.
The wider application of end-to-end learning methods to embodied decision-making domains remains bottlenecked by their reliance on a superabundance of training data representative of the target domain. Meta-reinforcement learning (meta-RL) approaches abandon the aim of zero-shot generalization--the goal of standard reinforcement learning (RL)--in favor of few-shot adaptation, and thus hold promise for bridging larger generalization gaps. While learning this meta-level adaptive behavior still requires substantial data, efficient environment simulators approaching real-world complexity are growing in prevalence. Even so, hand-designing sufficiently diverse and numerous simulated training tasks for these complex domains is prohibitively labor-intensive. Domain randomization (DR) and procedural generation (PG), offered as solutions to this problem, require simulators to possess carefully-defined parameters which directly translate to meaningful task diversity--a similarly prohibitive assumption. In this work, we present DIVA, an evolutionary approach for generating diverse training tasks in such complex, open-ended simulators. Like unsupervised environment design (UED) methods, DIVA can be applied to arbitrary parameterizations, but can additionally incorporate realistically-available domain knowledge--thus inheriting the flexibility and generality of UED, and the supervised structure embedded in well-designed simulators exploited by DR and PG. Our empirical results showcase DIVA's unique ability to overcome complex parameterizations and successfully train adaptive agent behavior, far outperforming competitive baselines from prior literature. These findings highlight the potential of such semi-supervised environment design (SSED) approaches, of which DIVA is the first humble constituent, to enable training in realistic simulated domains, and produce more robust and capable adaptive agents.
Recently, large language models (LLMs) have achieved significant progress in automated code generation. Despite their strong instruction-following capabilities, these models frequently struggled to align with user intent in coding scenarios. In particular, they were hampered by datasets that lacked diversity and failed to address specialized tasks or edge cases. Furthermore, challenges in supervised fine-tuning (SFT) and reinforcement learning from human feedback (RLHF) led to failures in generating precise, human-intent-aligned code. To tackle these challenges and improve the code generation performance for automated programming systems, we propose Feedback-driven Adaptive Long/short-term memory reinforced Coding Optimization (i.e., FALCON). FALCON is structured into two hierarchical levels. From the global level, long-term memory improves code quality by retaining and applying learned knowledge. At the local level, short-term memory allows for the incorporation of immediate feedback from compilers and AI systems. Additionally, we introduce meta-reinforcement learning with feedback rewards to solve the global-local bi-level optimization problem and enhance the model's adaptability across diverse code generation tasks. Extensive experiments demonstrate that our technique achieves state-of-the-art performance, leading other reinforcement learning methods by more than 4.5 percentage points on the MBPP benchmark and 6.1 percentage points on the Humaneval benchmark. The open-sourced code is publicly available at https://github.com/titurte/FALCON.
Effective decision-making in partially observable environments demands robust memory management. Despite their success in supervised learning, current deep-learning memory models struggle in reinforcement learning environments that are partially observable and long-term. They fail to efficiently capture relevant past information, adapt flexibly to changing observations, and maintain stable updates over long episodes. We theoretically analyze the limitations of existing memory models within a unified framework and introduce the Stable Hadamard Memory, a novel memory model for reinforcement learning agents. Our model dynamically adjusts memory by erasing no longer needed experiences and reinforcing crucial ones computationally efficiently. To this end, we leverage the Hadamard product for calibrating and updating memory, specifically designed to enhance memory capacity while mitigating numerical and learning challenges. Our approach significantly outperforms state-of-the-art memory-based methods on challenging partially observable benchmarks, such as meta-reinforcement learning, long-horizon credit assignment, and POPGym, demonstrating superior performance in handling long-term and evolving contexts.
Meta-reinforcement learning (Meta-RL) has attracted attention due to its capability to enhance reinforcement learning (RL) algorithms, in terms of data efficiency and generalizability. In this paper, we develop a bilevel optimization framework for meta-RL (BO-MRL) to learn the meta-prior for task-specific policy adaptation, which implements multiple-step policy optimization on one-time data collection. Beyond existing meta-RL analyses, we provide upper bounds of the expected optimality gap over the task distribution. This metric measures the distance of the policy adaptation from the learned meta-prior to the task-specific optimum, and quantifies the model's generalizability to the task distribution. We empirically validate the correctness of the derived upper bounds and demonstrate the superior effectiveness of the proposed algorithm over benchmarks.
As wireless networks grow to support more complex applications, the Open Radio Access Network (O-RAN) architecture, with its smart RAN Intelligent Controller (RIC) modules, becomes a crucial solution for real-time network data collection, analysis, and dynamic management of network resources including radio resource blocks and downlink power allocation. Utilizing artificial intelligence (AI) and machine learning (ML), O-RAN addresses the variable demands of modern networks with unprecedented efficiency and adaptability. Despite progress in using ML-based strategies for network optimization, challenges remain, particularly in the dynamic allocation of resources in unpredictable environments. This paper proposes a novel Meta Deep Reinforcement Learning (Meta-DRL) strategy, inspired by Model-Agnostic Meta-Learning (MAML), to advance resource block and downlink power allocation in O-RAN. Our approach leverages O-RAN's disaggregated architecture with virtual distributed units (DUs) and meta-DRL strategies, enabling adaptive and localized decision-making that significantly enhances network efficiency. By integrating meta-learning, our system quickly adapts to new network conditions, optimizing resource allocation in real-time. This results in a 19.8% improvement in network management performance over traditional methods, advancing the capabilities of next-generation wireless networks.
Monotonic Partially Observable Markov Decision Processes (POMDPs), where the system state progressively decreases until a restorative action is performed, can be used to model sequential repair problems effectively. This paper considers the problem of solving budget-constrained multi-component monotonic POMDPs, where a finite budget limits the maximal number of restorative actions. For a large number of components, solving such a POMDP using current methods is computationally intractable due to the exponential growth in the state space with an increasing number of components. To address this challenge, we propose a two-step approach. Since the individual components of a budget-constrained multi-component monotonic POMDP are only connected via the shared budget, we first approximate the optimal budget allocation among these components using an approximation of each component POMDP's optimal value function which is obtained through a random forest model. Subsequently, we introduce an oracle-guided meta-trained Proximal Policy Optimization (PPO) algorithm to solve each of the independent budget-constrained single-component monotonic POMDPs. The oracle policy is obtained by performing value iteration on the corresponding monotonic Markov Decision Process (MDP). This two-step method provides scalability in solving truly massive multi-component monotonic POMDPs. To demonstrate the efficacy of our approach, we consider a real-world maintenance scenario that involves inspection and repair of an administrative building by a team of agents within a maintenance budget. Finally, we perform a computational complexity analysis for a varying number of components to show the scalability of the proposed approach.
Training intelligent agents to navigate highly interactive environments presents significant challenges. While guided meta reinforcement learning (RL) approach that first trains a guiding policy to train the ego agent has proven effective in improving generalizability across scenarios with various levels of interaction, the state-of-the-art method tends to be overly sensitive to extreme cases, impairing the agents' performance in the more common scenarios. This study introduces a novel training framework that integrates guided meta RL with importance sampling (IS) to optimize training distributions iteratively for navigating highly interactive driving scenarios, such as T-intersections or roundabouts. Unlike traditional methods that may underrepresent critical interactions or overemphasize extreme cases during training, our approach strategically adjusts the training distribution towards more challenging driving behaviors using IS proposal distributions and applies the importance ratio to de-bias the result. By estimating a naturalistic distribution from real-world datasets and employing a mixture model for iterative training refinements, the framework ensures a balanced focus across common and extreme driving scenarios. Experiments conducted with both synthetic and naturalistic datasets demonstrate both accelerated training and performance improvements under highly interactive driving tasks.
Meta-Reinforcement Learning (Meta-RL) aims to acquire meta-knowledge for quick adaptation to diverse tasks. However, applying these policies in real-world environments presents a significant challenge in balancing rapid adaptability with adherence to environmental constraints. Our novel approach, Constraint Model Agnostic Meta Learning (C-MAML), merges meta learning with constrained optimization to address this challenge. C-MAML enables rapid and efficient task adaptation by incorporating task-specific constraints directly into its meta-algorithm framework during the training phase. This fusion results in safer initial parameters for learning new tasks. We demonstrate the effectiveness of C-MAML in simulated locomotion with wheeled robot tasks of varying complexity, highlighting its practicality and robustness in dynamic environments.
Fast adaptation to new tasks is extremely important for embodied agents in the real world. Meta-reinforcement learning (meta-RL) has emerged as an effective method to enable fast adaptation in unknown environments. Compared to on-policy meta-RL algorithms, off-policy algorithms rely heavily on efficient data sampling strategies to extract and represent the historical trajectories. However, little is known about how different data sampling methods impact the ability of meta-RL agents to represent unknown environments. Here, we investigate the impact of data sampling strategies on the exploration and adaptability of meta-RL agents. Specifically, we conducted experiments with two types of off-policy meta-RL algorithms based on Thompson sampling and Bayes-optimality theories in continuous control tasks within the MuJoCo environment and sparse reward navigation tasks. Our analysis revealed the long-memory and short-memory sequence sampling strategies affect the representation and adaptive capabilities of meta-RL agents. We found that the algorithm based on Bayes-optimality theory exhibited more robust and better adaptability than the algorithm based on Thompson sampling, highlighting the importance of appropriate data sampling strategies for the agent's representation of an unknown environment, especially in the case of sparse rewards.
Meta-Reinforcement Learning (Meta-RL) agents can struggle to operate across tasks with varying environmental features that require different optimal skills (i.e., different modes of behaviour). Using context encoders based on contrastive learning to enhance the generalisability of Meta-RL agents is now widely studied but faces challenges such as the requirement for a large sample size, also referred to as the $\log$-$K$ curse. To improve RL generalisation to different tasks, we first introduce Skill-aware Mutual Information (SaMI), an optimisation objective that aids in distinguishing context embeddings according to skills, thereby equipping RL agents with the ability to identify and execute different skills across tasks. We then propose Skill-aware Noise Contrastive Estimation (SaNCE), a $K$-sample estimator used to optimise the SaMI objective. We provide a framework for equipping an RL agent with SaNCE in practice and conduct experimental validation on modified MuJoCo and Panda-gym benchmarks. We empirically find that RL agents that learn by maximising SaMI achieve substantially improved zero-shot generalisation to unseen tasks. Additionally, the context encoder trained with SaNCE demonstrates greater robustness to a reduction in the number of available samples, thus possessing the potential to overcome the $\log$-$K$ curse.
Meta reinforcement learning sets a distribution over a set of tasks on which the agent can train at will, then is asked to learn an optimal policy for any test task efficiently. In this paper, we consider a finite set of tasks modeled through Markov decision processes with various dynamics. We assume to have endured a long training phase, from which the set of tasks is perfectly recovered, and we focus on regret minimization against the optimal policy in the unknown test task. Under a separation condition that states the existence of a state-action pair revealing a task against another, Chen et al. (2022) show that $O(M^2 \log(H))$ regret can be achieved, where $M, H$ are the number of tasks in the set and test episodes, respectively. In our first contribution, we demonstrate that the latter rate is nearly optimal by developing a novel lower bound for test-time regret minimization under separation, showing that a linear dependence with $M$ is unavoidable. Then, we present a family of stronger yet reasonable assumptions beyond separation, which we call strong identifiability, enabling algorithms achieving fast rates $\log (H)$ and sublinear dependence with $M$ simultaneously. Our paper provides a new understanding of the statistical barriers of test-time regret minimization and when fast rates can be achieved.
Meta-reinforcement learning has widely been used as a learning-to-learn framework to solve unseen tasks with limited experience. However, the aspect of constraint violations has not been adequately addressed in the existing works, making their application restricted in real-world settings. In this paper, we study the problem of meta-safe reinforcement learning (Meta-SRL) through the CMDP-within-online framework to establish the first provable guarantees in this important setting. We obtain task-averaged regret bounds for the reward maximization (optimality gap) and constraint violations using gradient-based meta-learning and show that the task-averaged optimality gap and constraint satisfaction improve with task-similarity in a static environment or task-relatedness in a dynamic environment. Several technical challenges arise when making this framework practical. To this end, we propose a meta-algorithm that performs inexact online learning on the upper bounds of within-task optimality gap and constraint violations estimated by off-policy stationary distribution corrections. Furthermore, we enable the learning rates to be adapted for every task and extend our approach to settings with a competing dynamically changing oracle. Finally, experiments are conducted to demonstrate the effectiveness of our approach.
This research delves deeply into Meta Reinforcement Learning (Meta RL) through a exploration focusing on defining generalization limits and ensuring convergence. By employing a approach this article introduces an innovative theoretical framework to meticulously assess the effectiveness and performance of Meta RL algorithms. We present an explanation of generalization limits measuring how well these algorithms can adapt to learning tasks while maintaining consistent results. Our analysis delves into the factors that impact the adaptability of Meta RL revealing the relationship, between algorithm design and task complexity. Additionally we establish convergence assurances by proving conditions under which Meta RL strategies are guaranteed to converge towards solutions. We examine the convergence behaviors of Meta RL algorithms across scenarios providing a comprehensive understanding of the driving forces behind their long term performance. This exploration covers both convergence and real time efficiency offering a perspective, on the capabilities of these algorithms.
Unmanned aerial vehicles (UAVs) with multiple antennas have recently been explored to improve capacity in wireless networks. However, the strict energy constraint of UAVs, given their simultaneous flying and communication tasks, renders the exploration of energy-efficient multi-antenna techniques indispensable for UAVs. Meanwhile, lens antenna subarray (LAS) emerges as a promising energy-efficient solution that has not been previously harnessed for this purpose. In this paper, we propose a LAS-aided multi-antenna UAV to serve ground users in the downlink transmission of the terahertz (THz) band, utilizing rate splitting multiple access (RSMA) for effective beam division multiplexing. We formulate an optimization problem of maximizing the total system spectral efficiency (SE). This involves optimizing the UAV's transmit beamforming and the common rate of RSMA. By recasting the optimization problem into a Markov decision process (MDP), we propose a deep deterministic policy gradient (DDPG)-based resource allocation mechanism tailored to capture problem dynamics and optimize its variables. Moreover, given the UAV's frequent mobility and consequential system reconfigurations, we fortify the trained DDPG model with a meta-learning strategy, enhancing its adaptability to system variations. Numerically, more than 20\% energy efficiency gain is achieved by our proposed LAS-aided multi-antenna UAV equipped with 4 lenses, compared to a single-lens UAV. Simulations also demonstrate that at a signal-to-noise (SNR) of 10 dB, the incorporation of RSMA results in a 22\% SE enhancement over conventional orthogonal beam division multiple access. Furthermore, the overall system SE improves by 27\%, when meta-learning is employed for fine-tuning the conventional DDPG method in literature.
This paper centers around a multiple-input-multiple-output (MIMO) visible light communication (VLC) system, where an unmanned aerial vehicle (UAV) benefits from a light emitting diode (LED) array to serve photo-diode (PD)-equipped users for illumination and communication simultaneously. Concerning the battery limitation of the UAV and considerable energy consumption of the LED array, a hybrid dimming control scheme is devised at the UAV that effectively controls the number of glared LEDs and thereby mitigates the overall energy consumption. To assess the performance of this system, a radio resource allocation problem is accordingly formulated for jointly optimizing the motion trajectory, transmit beamforming and LED selection at the UAV, assuming that channel state information (CSI) is partially available. By reformulating the optimization problem in Markov decision process (MDP) form, we propose a soft actor-critic (SAC) mechanism that captures the dynamics of the problem and optimizes its parameters. Additionally, regarding the high mobility of the UAV and thus remarkable rearrangement of the system, we enhance the trained SAC model by integrating a meta-learning strategy that enables more adaptation to system variations. By defining energy efficiency as a trade-off between the data rate and power consumption, simulations verify that upgrading a single-LED UAV by an array of 10 LEDs, exhibits 47% and 34% improvements in data rate and energy efficiency, albeit at the expense of 8% more power consumption.