Robots in uncertain real-world environments must perform both goal-directed and exploratory actions. However, most deep learning-based control methods neglect exploration and struggle under uncertainty. To address this, we adopt deep active inference, a framework that accounts for human goal-directed and exploratory actions. Yet, conventional deep active inference approaches face challenges due to limited environmental representation capacity and high computational cost in action selection. We propose a novel deep active inference framework that consists of a world model, an action model, and an abstract world model. The world model encodes environmental dynamics into hidden state representations at slow and fast timescales. The action model compresses action sequences into abstract actions using vector quantization, and the abstract world model predicts future slow states conditioned on the abstract action, enabling low-cost action selection. We evaluate the framework on object-manipulation tasks with a real-world robot. Results show that it achieves high success rates across diverse manipulation tasks and switches between goal-directed and exploratory actions in uncertain settings, while making action selection computationally tractable. These findings highlight the importance of modeling multiple timescale dynamics and abstracting actions and state transitions.

We study how the word usage of U.S. Congressional speeches varies across states and party affiliation, how words are used differently across sections of the ArXiv, and how the co-purchase patterns of groceries can vary across seasons.

With the recent success of world-model agents, which extend the core idea of model-based reinforcement learning by learning a differentiable model for sample-efficient control across diverse tasks, active inference (AIF) offers a complementary, neuroscience-grounded paradigm that unifies perception, learning, and action within a single probabilistic framework powered by a generative model. Despite this promise, practical AIF agents still rely on accurate immediate predictions and exhaustive planning, a limitation that is exacerbated in delayed environments requiring plans over long horizons, tens to hundreds of steps. Moreover, most existing agents are evaluated on robotic or vision benchmarks which, while natural for biological agents, fall short of real-world industrial complexity. We address these limitations with a generative-policy architecture featuring (i) a multi-step latent transition that lets the generative model predict an entire horizon in a single look-ahead, (ii) an integrated policy network that enables the transition and receives gradients of the expected free energy, (iii) an alternating optimization scheme that updates model and policy from a replay buffer, and (iv) a single gradient step that plans over long horizons, eliminating exhaustive planning from the control loop. We evaluate our agent in an environment that mimics a realistic industrial scenario with delayed and long-horizon settings. The empirical results confirm the effectiveness of the proposed approach, demonstrating the coupled world-model with the AIF formalism yields an end-to-end probabilistic controller capable of effective decision making in delayed, long-horizon settings without handcrafted rewards or expensive planning.

Handling heterogeneity and unpredictability are two core problems in pervasive computing. The challenge is to seamlessly integrate devices with varying computational resources in a dynamic environment to form a cohesive system that can fulfill the needs of all participants. Existing work on systems that adapt to changing requirements typically focuses on optimizing individual variables or low-level Service Level Objectives (SLOs), such as constraining the usage of specific resources. While low-level control mechanisms permit fine-grained control over a system, they introduce considerable complexity, particularly in dynamic environments. To this end, we propose drawing from Active Inference (AIF), a neuroscientific framework for designing adaptive agents. Specifically, we introduce a conceptual agent for heterogeneous pervasive systems that permits setting global systems constraints as high-level SLOs. Instead of manually setting low-level SLOs, the system finds an equilibrium that can adapt to environmental changes. We demonstrate the viability of AIF agents with an extensive experiment design, using heterogeneous and lifelong federated learning as an application scenario. We conduct our experiments on a physical testbed of devices with different resource types and vendor specifications. The results provide convincing evidence that an AIF agent can adapt a system to environmental changes. In particular, the AIF agent can balance competing SLOs in resource heterogeneous environments to ensure up to 98% fulfillment rate.

We study how the word usage of U.S. Congressional speeches varies across states and party affiliation, how words are used differently across sections of the ArXiv, and how the copurchase patterns of groceries can vary across seasons. Key to the success of our method is that the groups share statistical information. We develop two sharing strategies: hierarchical modeling and amortization. We demonstrate the benefits of this approach in empirical studies of speeches, abstracts, and shopping baskets.

Expected free energy (EFE) is a central quantity in active inference which has recently gained popularity due to its intuitive decomposition of the expected value of control into a pragmatic and an epistemic component. While numerous conjectures have been made to justify EFE as a decision making objective function, the most widely accepted is still its intuitiveness and resemblance to variational free energy in approximate Bayesian inference. In this work, we take a bottom up approach and ask: taking EFE as given, what's the resulting agent's optimality gap compared with a reward-driven reinforcement learning (RL) agent, which is well understood? By casting EFE under a particular class of belief MDP and using analysis tools from RL theory, we show that EFE approximates the Bayes optimal RL policy via information value. We discuss the implications for objective specification of active inference agents.

We investigate the application of active inference in developing energy-efficient control agents for manufacturing systems. Active inference, rooted in neuroscience, provides a unified probabilistic framework integrating perception, learning, and action, with inherent uncertainty quantification elements. Our study explores deep active inference, an emerging field that combines deep learning with the active inference decision-making framework. Leveraging a deep active inference agent, we focus on controlling parallel and identical machine workstations to enhance energy efficiency. We address challenges posed by the problem's stochastic nature and delayed policy response by introducing tailored enhancements to existing agent architectures. Specifically, we introduce multi-step transition and hybrid horizon methods to mitigate the need for complex planning. Our experimental results demonstrate the effectiveness of these enhancements and highlight the potential of the active inference-based approach.

For robotic decision-making under uncertainty, the balance between exploitation and exploration of available options must be carefully taken into account. In this study, we introduce a new variant of contextual multi-armed bandits called observation-augmented CMABs (OA-CMABs) wherein a decision-making agent can utilize extra outcome observations from an external information source. CMABs model the expected option outcomes as a function of context features and hidden parameters, which are inferred from previous option outcomes. In OA-CMABs, external observations are also a function of context features and thus provide additional evidence about the hidden parameters. Yet, if an external information source is error-prone, the resulting posterior updates can harm decision-making performance unless the presence of errors is considered. To this end, we propose a robust Bayesian inference process for OA-CMABs that is based on the concept of probabilistic data validation. Our approach handles complex mixture model parameter priors and hybrid observation likelihoods for semantic data sources, allowing us to develop validation algorithms based on recently develop probabilistic semantic data association techniques. Furthermore, to more effectively cope with the combined sources of uncertainty in OA-CMABs, we derive a new active inference algorithm for option selection based on expected free energy minimization. This generalizes previous work on active inference for bandit-based robotic decision-making by accounting for faulty observations and non-Gaussian inference. Our approaches are demonstrated on a simulated asynchronous search site selection problem for space exploration. The results show that even if incorrect observations are provided by external information sources, efficient decision-making and robust parameter inference are still achieved in a wide variety of experimental conditions.

Active inference is a theory of perception, learning and decision making, which can be applied to neuroscience, robotics, and machine learning. Recently, reasearch has been taking place to scale up this framework using Monte-Carlo tree search and deep learning. The goal of this activity is to solve more complicated tasks using deep active inference. First, we review the existing literature, then, we progresively build a deep active inference agent. For two agents, we have experimented with five definitions of the expected free energy and three different action selection strategies. According to our experiments, the models able to solve the dSprites environment are the ones that maximise rewards. Finally, we compare the similarity of the representation learned by the layers of various agents using centered kernel alignment. Importantly, the agent maximising reward and the agent minimising expected free energy learn very similar representations except for the last layer of the critic network (reflecting the difference in learning objective), and the variance layers of the transition and encoder networks. We found that the reward maximising agent is a lot more certain than the agent minimising expected free energy. This is because the agent minimising expected free energy always picks the action down, and does not gather enough data for the other actions. In contrast, the agent maximising reward, keeps on selecting the actions left and right, enabling it to successfully solve the task. The only difference between those two agents is the epistemic value, which aims to make the outputs of the transition and encoder networks as close as possible. Thus, the agent minimising expected free energy picks a single action (down), and becomes an expert at predicting the future when selecting this action. This makes the KL divergence between the output of the transition and encoder networks small.

We apply a new method for learning equations from data -- Exhaustive Symbolic Regression (ESR) -- to late-type galaxy dynamics as encapsulated in the radial acceleration relation (RAR). Relating the centripetal acceleration due to baryons, $g_\text{bar}$, to the total dynamical acceleration, $g_\text{obs}$, the RAR has been claimed to manifest a new law of nature due to its regularity and tightness, in agreement with Modified Newtonian Dynamics (MOND). Fits to this relation have been restricted by prior expectations to particular functional forms, while ESR affords an exhaustive and nearly prior-free search through functional parameter space to identify the equations optimally trading accuracy with simplicity. Working with the SPARC data, we find the best functions typically satisfy $g_\text{obs} \propto g_\text{bar}$ at high $g_\text{bar}$, although the coefficient of proportionality is not clearly unity and the deep-MOND limit $g_\text{obs} \propto \sqrt{g_\text{bar}}$ as $g_\text{bar} \to 0$ is little evident at all. By generating mock data according to MOND with or without the external field effect, we find that symbolic regression would not be expected to identify the generating function or reconstruct successfully the asymptotic slopes. We conclude that the limited dynamical range and significant uncertainties of the SPARC RAR preclude a definitive statement of its functional form, and hence that this data alone can neither demonstrate nor rule out law-like gravitational behaviour.