Canonical neural networks perform active inference

Active Inference for Learning and Development in Embodied Neuromorphic Agents

Taking inspiration from humans can help catalyse embodied AI solutions for important real-world applications. Current human-inspired tools include neuromorphic systems and the developmental approach to learning. However, this developmental neurorobotics approach is currently lacking important frameworks for human-like computation and learning. We propose that human-like computation is inherently embodied, with its interface to the world being neuromorphic, and its learning processes operating across different timescales. These constraints necessitate a unified framework: active inference, underpinned by the free energy principle (FEP). Herein, we describe theoretical and empirical support for leveraging this framework in embodied neuromorphic agents with autonomous mental development. We additionally outline current implementation approaches (including toolboxes) and challenges, and we provide suggestions for next steps to catalyse this important field.

The memory systems of the human brain and generative artificial intelligence

Generative Artificial Intelligence foundation models (for example Generative Pre-trained Transformer - GPT - models) can generate the next token given a sequence of tokens. How can this 'generative AI' be compared with the 'real' intelligence of the human brain, when for example a human generates a whole memory in response to an incomplete retrieval cue, and then generates further prospective thoughts? Here these two types of generative intelligence, artificial in machines and real in the human brain are compared, and it is shown how when whole memories are generated by hippocampal recall in response to an incomplete retrieval cue, what the human brain computes, and how it computes it, are very different from generative AI. Key differences are the use of local associative learning rules in the hippocampal memory system, and of non-local backpropagation of error learning in AI. Indeed, it is argued that the whole operation of the human brain is performed computationally very differently to what is implemented in generative AI. Moreover, it is emphasized that the primate including human hippocampal system includes computations about spatial view and where objects and people are in scenes, whereas in rodents the emphasis is on place cells and path integration by movements between places. This comparison with generative memory and processing in the human brain has interesting implications for the further development of generative AI and for neuroscience research.

On Predictive Planning and Counterfactual Learning in Active Inference

Given the rapid advancement of artificial intelligence, understanding the foundations of intelligent behaviour is increasingly important. Active inference, regarded as a general theory of behaviour, offers a principled approach to probing the basis of sophistication in planning and decision-making. This paper examines two decision-making schemes in active inference based on "planning" and "learning from experience". Furthermore, we also introduce a mixed model that navigates the data complexity trade-off between these strategies, leveraging the strengths of both to facilitate balanced decision-making. We evaluate our proposed model in a challenging grid-world scenario that requires adaptability from the agent. Additionally, our model provides the opportunity to analyse the evolution of various parameters, offering valuable insights and contributing to an explainable framework for intelligent decision-making.

Neuron-level Prediction and Noise can Implement Flexible Reward-Seeking Behavior

We show that neural networks can implement reward-seeking behavior using only local predictive updates and internal noise. These networks are capable of autonomous interaction with an environment and can switch between explore and exploit behavior, which we show is governed by attractor dynamics. Networks can adapt to changes in their architectures, environments, or motor interfaces without any external control signals. When networks have a choice between different tasks, they can form preferences that depend on patterns of noise and initialization, and we show that these preferences can be biased by network architectures or by changing learning rates. Our algorithm presents a flexible, biologically plausible way of interacting with environments without requiring an explicit environmental reward function, allowing for behavior that is both highly adaptable and autonomous. Code is available at https://github.com/ccli3896/PaN.

The empirical status of predictive coding and active inference

Research on predictive processing models has focused largely on two specific algorithmic theories: Predictive Coding for perception and Active Inference for decision-making. While these interconnected theories possess broad explanatory potential, they have only recently begun to receive direct empirical evaluation. Here, we review recent studies of Predictive Coding and Active Inference with a focus on evaluating the degree to which they are empirically supported. For Predictive Coding, we find that existing empirical evidence offers modest support. However, some positive results can also be explained by alternative feedforward (e.g., feature detection-based) models. For Active Inference, most empirical studies have focused on fitting these models to behavior as a means of identifying and explaining individual or group differences. While Active Inference models tend to explain behavioral data reasonably well, there has not been a focus on testing empirical validity of active inference theory per se, which would require formal comparison to other models (e.g., non-Bayesian or model-free reinforcement learning models). This review suggests that, while promising, a number of specific research directions are still necessary to evaluate the empirical adequacy and explanatory power of these algorithms.

Fetal brain activity and the free energy principle

OBJECTIVES

To study whether the free energy principle can explain fetal brain activity and the existence of fetal consciousness via a chaotic dimension derived using artificial intelligence.

METHODS

In this observational study, we used a four-dimensional ultrasound technique obtained to collect images of fetal faces from pregnancies at 27-37 weeks of gestation, between February and December 2021. We developed an artificial intelligence classifier that recognizes fetal facial expressions, which are thought to relate to fetal brain activity. We then applied the classifier to video files of facial images to generate each expression category's probabilities. We calculated the chaotic dimensions from the probability lists, and we created and investigated the free energy principle's mathematical model that was assumed to be linked to the chaotic dimension. We used a Mann-Whitney test, linear regression test, and one-way analysis of variance for statistical analysis.

RESULTS

The chaotic dimension revealed that the fetus had dense and sparse states of brain activity, which fluctuated at a statistically significant level. The chaotic dimension and free energy were larger in the sparse state than in the dense state.

CONCLUSIONS

The fluctuating free energy suggests consciousness seemed to exist in the fetus after 27 weeks.

The computational power of the human brain

At the end of the 20th century, analog systems in computer science have been widely replaced by digital systems due to their higher computing power. Nevertheless, the question keeps being intriguing until now: is the brain analog or digital? Initially, the latter has been favored, considering it as a Turing machine that works like a digital computer. However, more recently, digital and analog processes have been combined to implant human behavior in robots, endowing them with artificial intelligence (AI). Therefore, we think it is timely to compare mathematical models with the biology of computation in the brain. To this end, digital and analog processes clearly identified in cellular and molecular interactions in the Central Nervous System are highlighted. But above that, we try to pinpoint reasons distinguishing in silico computation from salient features of biological computation. First, genuinely analog information processing has been observed in electrical synapses and through gap junctions, the latter both in neurons and astrocytes. Apparently opposed to that, neuronal action potentials (APs) or spikes represent clearly digital events, like the yes/no or 1/0 of a Turing machine. However, spikes are rarely uniform, but can vary in amplitude and widths, which has significant, differential effects on transmitter release at the presynaptic terminal, where notwithstanding the quantal (vesicular) release itself is digital. Conversely, at the dendritic site of the postsynaptic neuron, there are numerous analog events of computation. Moreover, synaptic transmission of information is not only neuronal, but heavily influenced by astrocytes tightly ensheathing the majority of synapses in brain (tripartite synapse). At least at this point, LTP and LTD modifying synaptic plasticity and believed to induce short and long-term memory processes including consolidation (equivalent to RAM and ROM in electronic devices) have to be discussed. The present knowledge of how the brain stores and retrieves memories includes a variety of options (e.g., neuronal network oscillations, engram cells, astrocytic syncytium). Also epigenetic features play crucial roles in memory formation and its consolidation, which necessarily guides to molecular events like gene transcription and translation. In conclusion, brain computation is not only digital or analog, or a combination of both, but encompasses features in parallel, and of higher orders of complexity.

Experimental validation of the free-energy principle with in vitro neural networks

Empirical applications of the free-energy principle are not straightforward because they entail a commitment to a particular process theory, especially at the cellular and synaptic levels. Using a recently established reverse engineering technique, we confirm the quantitative predictions of the free-energy principle using in vitro networks of rat cortical neurons that perform causal inference. Upon receiving electrical stimuli-generated by mixing two hidden sources-neurons self-organised to selectively encode the two sources. Pharmacological up- and downregulation of network excitability disrupted the ensuing inference, consistent with changes in prior beliefs about hidden sources. As predicted, changes in effective synaptic connectivity reduced variational free energy, where the connection strengths encoded parameters of the generative model. In short, we show that variational free energy minimisation can quantitatively predict the self-organisation of neuronal networks, in terms of their responses and plasticity. These results demonstrate the applicability of the free-energy principle to in vitro neural networks and establish its predictive validity in this setting.

Distinctive properties of biological neural networks and recent advances in bottom-up approaches toward a better biologically plausible neural network

Although it may appear infeasible and impractical, building artificial intelligence (AI) using a bottom-up approach based on the understanding of neuroscience is straightforward. The lack of a generalized governing principle for biological neural networks (BNNs) forces us to address this problem by converting piecemeal information on the diverse features of neurons, synapses, and neural circuits into AI. In this review, we described recent attempts to build a biologically plausible neural network by following neuroscientifically similar strategies of neural network optimization or by implanting the outcome of the optimization, such as the properties of single computational units and the characteristics of the network architecture. In addition, we proposed a formalism of the relationship between the set of objectives that neural networks attempt to achieve, and neural network classes categorized by how closely their architectural features resemble those of BNN. This formalism is expected to define the potential roles of top-down and bottom-up approaches for building a biologically plausible neural network and offer a map helping the navigation of the gap between neuroscience and AI engineering.

Free energy and inference in living systems

Organisms are non-equilibrium, stationary systems self-organized via spontaneous symmetry breaking and undergoing metabolic cycles with broken detailed balance in the environment. The thermodynamic free-energy (FE) principle describes an organism's homeostasis as the regulation of biochemical work constrained by the physical FE cost. By contrast, recent research in neuroscience and theoretical biology explains a higher organism's homeostasis and allostasis as Bayesian inference facilitated by the informational FE. As an integrated approach to living systems, this study presents an FE minimization theory overarching the essential features of both the thermodynamic and neuroscientific FE principles. Our results reveal that the perception and action of animals result from active inference entailed by FE minimization in the brain, and the brain operates as a Schrödinger's machine conducting the neural mechanics of minimizing sensory uncertainty. A parsimonious model suggests that the Bayesian brain develops the optimal trajectories in neural manifolds and induces a dynamic bifurcation between neural attractors in the process of active inference.

Hierarchical predictive coding in distributed pain circuits

Predictive coding is a computational theory on describing how the brain perceives and acts, which has been widely adopted in sensory processing and motor control. Nociceptive and pain processing involves a large and distributed network of circuits. However, it is still unknown whether this distributed network is completely decentralized or requires networkwide coordination. Multiple lines of evidence from human and animal studies have suggested that the cingulate cortex and insula cortex (cingulate-insula network) are two major hubs in mediating information from sensory afferents and spinothalamic inputs, whereas subregions of cingulate and insula cortices have distinct projections and functional roles. In this mini-review, we propose an updated hierarchical predictive coding framework for pain perception and discuss its related computational, algorithmic, and implementation issues. We suggest active inference as a generalized predictive coding algorithm, and hierarchically organized traveling waves of independent neural oscillations as a plausible brain mechanism to integrate bottom-up and top-down information across distributed pain circuits.

The Emperor Is Naked: Replies to commentaries on the target article

The 35 commentaries cover a wide range of topics and take many different stances on the issues explored by the target article. We have organised our response to the commentaries around three central questions: Are Friston blankets just Pearl blankets? What ontological and metaphysical commitments are implied by the use of Friston blankets? What kind of explanatory work are Friston blankets capable of? We conclude our reply with a short critical reflection on the indiscriminate use of both Markov blankets and the free energy principle.

Active inference leads to Bayesian neurophysiology

The neuronal substrates that implement the free-energy principle and ensuing active inference at the neuron and synapse level have not been fully elucidated. This Review considers possible neuronal substrates underlying the principle. First, the foundations of the free-energy principle are introduced, and then its ability to empirically explain various brain functions and psychological and biological phenomena in terms of Bayesian inference is described. Mathematically, the dynamics of neural activity and plasticity that minimise a cost function can be cast as performing Bayesian inference that minimises variational free energy. This equivalence licenses the adoption of the free-energy principle as a universal characterisation of neural networks. Further, the neural network structure itself represents a generative model under which an agent operates. A virtue of this perspective is that it enables the formal association of neural network properties with prior beliefs that regulate inference and learning. The possible neuronal substrates that implement prior beliefs and how to empirically examine the theory are discussed. This perspective renders brain activity explainable, leading to a deeper understanding of the neuronal mechanisms underlying basic psychology and psychiatric disorders in terms of an implicit generative model.