Predictive coding networks are neuroscience-inspired models with roots in both Bayesian statistics and neuroscience. Training such models, however, is quite inefficient and unstable. In this work, we show how by simply changing the temporal scheduling of the update rule for the synaptic weights leads to an algorithm that is much more efficient and stable than the original one, and has theoretical guarantees in terms of convergence. The proposed algorithm, that we call incremental predictive coding (iPC) is also more biologically plausible than the original one, as it it fully automatic. In an extensive set of experiments, we show that iPC constantly performs better than the original formulation on a large number of benchmarks for image classification, as well as for the training of both conditional and masked language models, in terms of test accuracy, efficiency, and convergence with respect to a large set of hyperparameters.

Animals can adapt their preferences for different types of reward according to physiological state, such as hunger or thirst. To explain this ability, we employ a simple multi-objective reinforcement learning model that learns multiple values according to different reward dimensions such as food or water. We show that by weighting these learned values according to the current needs, behaviour may be flexibly adapted to present preferences. This model predicts that individual dopamine neurons should encode the errors associated with some reward dimensions more than with others. To provide a preliminary test of this prediction, we reanalysed a small dataset obtained from a single primate in an experiment which to our knowledge is the only published study where the responses of dopamine neurons to stimuli predicting distinct types of rewards were recorded. We observed that in addition to subjective economic value, dopamine neurons encode a gradient of reward dimensions; some neurons respond most to stimuli predicting food rewards while the others respond more to stimuli predicting fluids. We also proposed a possible implementation of the model in the basal ganglia network, and demonstrated how the striatal system can learn values in multiple dimensions, even when dopamine neurons encode mixtures of prediction error from different dimensions. Additionally, the model reproduces the instant generalisation to new physiological states seen in dopamine responses and in behaviour. Our results demonstrate how a simple neural circuit can flexibly guide behaviour according to animals' needs.

An autoassociative memory model is a function that, given a set of data points, takes as input an arbitrary vector and outputs the most similar data point from the memorized set. However, popular memory models fail to retrieve images even when the corruption is mild and easy to detect for a human evaluator. This is because similarities are evaluated in the raw pixel space, which does not contain any semantic information about the images. This problem can be easily solved by computing similarities in an embedding space instead of the pixel space. We show that an effective way of computing such embeddings is via a network pretrained with a contrastive loss. As the dimension of embedding spaces is often significantly smaller than the pixel space, we also have a faster computation of similarity scores. We test this method on complex datasets such as CIFAR10 and STL10. An additional drawback of current models is the need of storing the whole dataset in the pixel space, which is often extremely large. We relax this condition and propose a class of memory models that only stores low-dimensional semantic embeddings, and uses them to retrieve similar, but not identical, memories. We demonstrate a proof of concept of this method on a simple task on the MNIST dataset.

The computational principles adopted by the hippocampus in associative memory (AM) tasks have been one of the most studied topics in computational and theoretical neuroscience. Recent theories suggested that AM and the predictive activities of the hippocampus could be described within a unitary account, and that predictive coding underlies the computations supporting AM in the hippocampus. Following this theory, a computational model based on classical hierarchical predictive networks was proposed and was shown to perform well in various AM tasks. However, this fully hierarchical model did not incorporate recurrent connections, an architectural component of the CA3 region of the hippocampus that is crucial for AM. This makes the structure of the model inconsistent with the known connectivity of CA3 and classical recurrent models such as Hopfield Networks, which learn the covariance of inputs through their recurrent connections to perform AM. Earlier PC models that learn the covariance information of inputs explicitly via recurrent connections seem to be a solution to these issues. Here, we show that although these models can perform AM, they do it in an implausible and numerically unstable way. Instead, we propose alternatives to these earlier covariance-learning predictive coding networks, which learn the covariance information implicitly and plausibly, and can use dendritic structures to encode prediction errors. We show analytically that our proposed models are perfectly equivalent to the earlier predictive coding model learning covariance explicitly, and encounter no numerical issues when performing AM tasks in practice. We further show that our models can be combined with hierarchical predictive coding networks to model the hippocampo-neocortical interactions. Our models provide a biologically plausible approach to modelling the hippocampal network, pointing to a potential computational mechanism during hippocampal memory formation and recall, which employs both predictive coding and covariance learning based on the recurrent network structure of the hippocampus.

The automatic subpolicy discovery approach in hierarchical reinforcement learning (HRL) has recently achieved promising performance on sparse reward tasks. This accelerates transfer learning and unsupervised intelligent creatures while eliminating the domain-specific knowledge constraint. Most previously developed approaches are demonstrated to suffer from collapsing into the situation where one subpolicy dominates the whole task, since they cannot ensure the diversity of different subpolicies. In contrast, this article proposes a task-agnostic regularizer (TAR) for learning diverse subpolicies in HRL. Specifically, we first formulate the discovery of diverse subpolicies as a trajectory inference problem and then propose a corresponding information-theoretic objective to encourage diversity. Subsequently, considering computability, we instantiate the objective as two simplifications for discrete and continuous action spaces. We extensively evaluate the proposed diversity-driven regularizer on three HRL task domains: 1) meta reinforcement learning; 2) hierarchical policy learning in the option framework; and 3) unsupervised subpolicy discovery. The extensive results obtained show that our TAR approach can improve upon the state-of-the-art performance on all three HRL domains without modifying any existing hyperparameters, indicating the wide applicability and robustness of our approach.

Deep learning has redefined AI thanks to the rise of artificial neural networks, which are inspired by neuronal networks in the brain. Through the years, these interactions between AI and neuroscience have brought immense benefits to both fields, allowing neural networks to be used in a plethora of applications. Neural networks use an efficient implementation of reverse differentiation, called backpropagation (BP). This algorithm, however, is often criticized for its biological implausibility (e.g., lack of local update rules for the parameters). Therefore, biologically plausible learning methods that rely on predictive coding (PC), a framework for describing information processing in the brain, are increasingly studied. Recent works prove that these methods can approximate BP up to a certain margin on multilayer perceptrons (MLPs), and asymptotically on any other complex model, and that zerodivergence inference learning (Z-IL), a variant of PC, is able to exactly implement BP on MLPs. However, the recent literature shows also that there is no biologically plausible method yet that can exactly replicate the weight update of BP on complex models. To fill this gap, in this paper, we generalize (PC and) Z-IL by directly defining it on computational graphs, and show that it can perform exact reverse differentiation. What results is the first PC (and so biologically plausible) algorithm that is equivalent to BP in the way of updating parameters on any neural network, providing a bridge between the interdisciplinary research of neuroscience and deep learning. Furthermore, the above results in particular also immediately provide a novel local and parallel implementation of BP.

A large number of neural network models of associative memory have been proposed in the literature. These include the classical Hopfield networks (HNs), sparse distributed memories (SDMs), and more recently the modern continuous Hopfield networks (MCHNs), which possess close links with self-attention in machine learning. In this paper, we propose a general framework for understanding the operation of such memory networks as a sequence of three operations: similarity, separation, and projection. We derive all these memory models as instances of our general framework with differing similarity and separation functions. We extend the mathematical framework of Krotov et al (2020) to express general associative memory models using neural network dynamics with local computation, and derive a general energy function that is a Lyapunov function of the dynamics. Finally, using our framework, we empirically investigate the capacity of using different similarity functions for these associative memory models, beyond the dot product similarity measure, and demonstrate empirically that Euclidean or Manhattan distance similarity metrics perform substantially better in practice on many tasks, enabling a more robust retrieval and higher memory capacity than existing models.

The backpropagation of error algorithm (BP) used to train deep neural networks has been fundamental to the successes of deep learning. However, it requires sequential backwards updates and non-local computations which make it challenging to parallelize at scale and is unlike how learning works in the brain. Neuroscience-inspired learning algorithms, however, such as \emph{predictive coding} which utilize local learning have the potential to overcome these limitations and advance beyond deep learning technologies in the future. While predictive coding originated in theoretical neuroscience as a model of information processing in the cortex, recent work has developed the idea into a general-purpose algorithm able to train neural networks using only local computations. In this survey, we review works that have contributed to this perspective and demonstrate the close connection between predictive coding and backpropagation in terms of generalization quality, as well as works that highlight the multiple advantages of using predictive coding models over backprop-trained neural networks. Specifically, we show the substantially greater flexibility of predictive coding networks against equivalent deep neural networks, which can function as classifiers, generators, and associative memories simultaneously, and can be defined on arbitrary graph topologies. Finally, we review direct benchmarks of predictive coding networks on machine learning classification tasks, as well as its close connections to control theory and applications in robotics.

A large amount of recent research has the far-reaching goal of finding training methods for deep neural networks that can serve as alternatives to backpropagation~(BP). A prominent example is predictive coding (PC), which is a neuroscience-inspired method that performs inference on hierarchical Gaussian generative models. These methods, however, fail to keep up with modern neural networks, as they are unable to replicate the dynamics of complex layers and activation functions. In this work, we solve this problem by generalizing PC to arbitrary probability distributions, enabling the training of architectures, such as transformers, that are hard to approximate with only Gaussian assumptions. We perform three experimental analyses. First, we study the gap between our method and the standard formulation of PC on multiple toy examples. Second, we test the reconstruction quality on variational autoencoders, where our method reaches the same reconstruction quality as BP. Third, we show that our method allows us to train transformer networks and achieve performance comparable with BP on conditional language models. More broadly, this method allows neuroscience-inspired learning to be applied to multiple domains, since the internal distributions can be flexibly adapted to the data, tasks, and architectures used.

Training with backpropagation (BP) in standard deep learning consists of two main steps: a forward pass that maps a data point to its prediction, and a backward pass that propagates the error of this prediction back through the network. This process is highly effective when the goal is to minimize a specific objective function. However, it does not allow training on networks with cyclic or backward connections. This is an obstacle to reaching brain-like capabilities, as the highly complex heterarchical structure of the neural connections in the neocortex are potentially fundamental for its effectiveness. In this paper, we show how predictive coding (PC), a theory of information processing in the cortex, can be used to perform inference and learning on arbitrary graph topologies. We experimentally show how this formulation, called PC graphs, can be used to flexibly perform different tasks with the same network by simply stimulating specific neurons. This enables the model to be queried on stimuli with different structures, such as partial images, images with labels, or images without labels. We conclude by investigating how the topology of the graph influences the final performance, and comparing against simple baselines trained with BP.

The great success of deep learning has boosted the fast development of video quality enhancement. However, existing methods mainly focus on enhancing the objective quality of compressed video, and ignore their perceptual quality that plays a key role in determining quality of experience (QoE) of videos. In this paper, we aim at enhancing the perceptual quality of compressed video. Our main observation is that perceptual quality enhancement mostly relies on recovering the high-frequency details with fine textures. Accordingly, we propose a novel generative adversarial network (GAN) based on multi-level wavelet packet transform (WPT), which is called multi-level wavelet-based GAN+ (MW-GAN+), to exploit high-frequency details for enhancing the perceptual quality of compressed video. In MW-GAN+, we first propose a multi-level wavelet pixel-adaptive (MWP) module to extract temporal information across video frames, such that frame similarity can be utilized in recovering high-frequency details. Then, a wavelet reconstruction network, consisting of wavelet-dense residual blocks (WDRB), is developed to recover high-frequency details in a multi-level manner for enhanced frame reconstruction. Finally, we develop a 3D discriminator to encourage temporal coherence with a 3D-CNN based architecture. Experimental results demonstrate the superiority of our method over state-of-the-art methods in enhancing the perceptual quality of compressed video. Our code is available athttps://github.com/IceClear/MW-GAN.

The omnidirectional images (ODIs) are usually at low-resolution, due to the constraints of collection, storage and transmission. The traditional two-dimensional (2D) image super-resolution methods are not effective for spherical ODIs, because ODIs tend to have non-uniformly distributed pixel density and varying texture complexity across latitudes. In this work, we propose a novel latitude adaptive upscaling network (LAU-Net) for ODI super-resolution, which allows pixels at different latitudes to adopt distinct upscaling factors. Specifically, we introduce a Laplacian multi-level separation architecture to split an ODI into different latitude bands, and hierarchically upscale them with different factors. In addition, we propose a deep reinforcement learning scheme with a latitude adaptive reward, in order to automatically select optimal upscaling factors for different latitude bands. To the best of our knowledge, LAU-Net is the first attempt to consider the latitude difference for ODI super-resolution. Extensive results demonstrate that our LAU-Net significantly advances the super-resolution performance for ODIs. Codes are available at https://github.com/wangh-allen/LAU-Net.

Associative memories in the brain receive and store patterns of activity registered by the sensory neurons, and are able to retrieve them when necessary. Due to their importance in human intelligence, computational models of associative memories have been developed for several decades now. In this paper, we present a novel neural model for realizing associative memories, which is based on a hierarchical generative network that receives external stimuli via sensory neurons. It is trained using predictive coding, an error-based learning algorithm inspired by information processing in the cortex. To test the model's capabilities, we perform multiple retrieval experiments from both corrupted and incomplete data points. In an extensive comparison, we show that this new model outperforms in retrieval accuracy and robustness popular associative memory models, such as autoencoders trained via backpropagation, and modern Hopfield networks. In particular, in completing partial data points, our model achieves remarkable results on natural image datasets, such as ImageNet, with a surprisingly high accuracy, even when only a tiny fraction of pixels of the original images is presented. Our model provides a plausible framework to study learning and retrieval of memories in the brain, as it closely mimics the behavior of the hippocampus as a memory index and generative model.

Virtual cinematography refers to automatically selecting a natural-looking normal field-of-view (NFOV) from an entire 360$^{\circ}$video. In fact, virtual cinematography can be modeled as a deep reinforcement learning (DRL) problem, in which an agent makes actions related to NFOV selection according to the environment of 360$^{\circ}$video frames. More importantly, we find from our data analysis that the selected NFOVs attract significantly more attention than other regions, i.e., the NFOVs have high saliency. Therefore, in this paper, we propose an attention-based DRL (A-DRL) approach for virtual cinematography in 360$^{\circ}$video. Specifically, we develop a new DRL framework for automatic NFOV selection with the input of both the content, and saliency map of each 360$^{\circ}$frame. Then, we propose a new reward function for the DRL framework in our approach, which considers the saliency values, ground-truth, and smooth transition for NFOV selection. Subsequently, a simplified DenseNet (called Mini-DenseNet) is designed to learn the optimal policy via maximizing the reward. Based on the learned policy, the actions of NFOV can be made in our A-DRL approach for virtual cinematography of 360$^{\circ}$video. Extensive experiments show that our A-DRL approach outperforms other state-of-the-art virtual cinematography methods, over the datasets of Sports-360 video, and Pano2Vid.

Learning agents that are not only capable of taking tests, but also innovating is becoming a hot topic in AI. One of the most promising paths towards this vision is multi-agent learning, where agents act as the environment for each other, and improving each agent means proposing new problems for others. However, existing evaluation platforms are either not compatible with multi-agent settings, or limited to a specific game. That is, there is not yet a general evaluation platform for research on multi-agent intelligence. To this end, we introduce Arena, a general evaluation platform for multi-agent intelligence with 35 games of diverse logics and representations. Furthermore, multi-agent intelligence is still at the stage where many problems remain unexplored. Therefore, we provide a building toolkit for researchers to easily invent and build novel multi-agent problems from the provided game set based on a GUI-configurable social tree and five basic multi-agent reward schemes. Finally, we provide Python implementations of five state-of-the-art deep multi-agent reinforcement learning baselines. Along with the baseline implementations, we release a set of 100 best agents/teams that we can train with different training schemes for each game, as the base for evaluating agents with population performance. As such, the research community can perform comparisons under a stable and uniform standard. All the implementations and accompanied tutorials have been open-sourced for the community at https://sites.google.com/view/arena-unity/.

Intrinsic rewards were introduced to simulate how human intelligence works; they are usually evaluated by intrinsically-motivated play, i.e., playing games without extrinsic rewards but evaluated with extrinsic rewards. However, none of the existing intrinsic reward approaches can achieve human-level performance under this very challenging setting of intrinsically-motivated play. In this work, we propose a novel megalomania-driven intrinsic reward (called mega-reward), which, to our knowledge, is the first approach that achieves human-level performance in intrinsically-motivated play. Intuitively, mega-reward comes from the observation that infants' intelligence develops when they try to gain more control on entities in an environment; therefore, mega-reward aims to maximize the control capabilities of agents on given entities in a given environment. To formalize mega-reward, a relational transition model is proposed to bridge the gaps between direct and latent control. Experimental studies show that mega-reward (i) can greatly outperform all state-of-the-art intrinsic reward approaches, (ii) generally achieves the same level of performance as Ex-PPO and professional human-level scores, and (iii) has also a superior performance when it is incorporated with extrinsic rewards.

Automatic sub-policy discovery has recently received much attention in hierarchical reinforcement learning (HRL). The conventional approaches to learning sub-policies suffer from collapsing into just one sub-policy dominating the whole task, lacking techniques to ensure the diversity of different subpolicies. In this paper, we formulate the discovery of diverse sub-policies as a trajectory inference. Then, we propose an information-theoretic objective based on action distributions to encourage diversity. Moreover, two simplifications are derived on discrete and continuous action space for reducing the computation. Finally, the experimental results show that the proposed approach can further improve the state-of-theart approaches without modifying existing hyperparameters on two different HRL domains, suggesting the wide applicability and robustness of our approach.

Backpropagation (BP) has been the most successful algorithm used to train artificial neural networks. However, there are several gaps between BP and learning in biologically plausible neuronal networks of the brain (learning in the brain, or simply BL, for short), in particular, (1) it has been unclear to date, if BP can be implemented exactly via BL, (2) there is a lack of local plasticity in BP, i.e., weight updates require information that is not locally available, while BL utilizes only locally available information, and (3)~there is a lack of autonomy in BP, i.e., some external control over the neural network is required (e.g., switching between prediction and learning stages requires changes to dynamics and synaptic plasticity rules), while BL works fully autonomously. Bridging such gaps, i.e., understanding how BP can be approximated by BL, has been of major interest in both neuroscience and machine learning. Despite tremendous efforts, however, no previous model has bridged the gaps at a degree of demonstrating an equivalence to BP, instead, only approximations to BP have been shown. Here, we present for the first time a framework within BL that bridges the above crucial gaps. We propose a BL model that (1) produces \emph{exactly the same} updates of the neural weights as~BP, while (2)~employing local plasticity, i.e., all neurons perform only local computations, done simultaneously. We then modify it to an alternative BL model that (3) also works fully autonomously. Overall, our work provides important evidence for the debate on the long-disputed question whether the brain can perform~BP.

Hierarchical reinforcement learning (HRL) has recently shown promising advances on speeding up learning, improving the exploration, and discovering intertask transferable skills. Most recent works focus on HRL with two levels, i.e., a master policy manipulates subpolicies, which in turn manipulate primitive actions. However, HRL with multiple levels is usually needed in many real-world scenarios, whose ultimate goals are highly abstract, while their actions are very primitive. Therefore, in this paper, we propose a diversitydriven extensible HRL (DEHRL), where an extensible and scalable framework is built and learned levelwise to realize HRL with multiple levels. DEHRL follows a popular assumption: diverse subpolicies are useful, i.e., subpolicies are believed to be more useful if they are more diverse. However, existing implementations of this diversity assumption usually have their own drawbacks, which makes them inapplicable to HRL with multiple levels. Consequently, we further propose a novel diversity-driven solution to achieve this assumption in DEHRL. Experimental studies evaluate DEHRL with nine baselines from four perspectives in two domains; the results show that DEHRL outperforms the state-of-the-art baselines in all four aspects.

Panoramic video provides immersive and interactive experience by enabling humans to control the field of view (FoV) through head movement (HM). Thus, HM plays a key role in modeling human attention on panoramic video. This paper establishes a database collecting subjects' HM in panoramic video sequences. From this database, we find that the HM data are highly consistent across subjects. Furthermore, we find that deep reinforcement learning (DRL) can be applied to predict HM positions, via maximizing the reward of imitating human HM scanpaths through the agent's actions. Based on our findings, we propose a DRL-based HM prediction (DHP) approach with offline and online versions, called offline-DHP and online-DHP. In offline-DHP, multiple DRL workflows are run to determine potential HM positions at each panoramic frame. Then, a heat map of the potential HM positions, named the HM map, is generated as the output of offline-DHP. In online-DHP, the next HM position of one subject is estimated given the currently observed HM position, which is achieved by developing a DRL algorithm upon the learned offline-DHP model. Finally, the experiments validate that our approach is effective in both offline and online prediction of HM positions for panoramic video, and that the learned offline-DHP model can improve the performance of online-DHP.