Cognitive Architecture Research Articles

Cortex level connectivity between ACT-R modules during EEG-based n-back task

AbstractFinding the synchronization between Electroencephalography (EEG) and human cognition is an essential aspect of cognitive neuroscience. Adaptive Control of Thought-Rational (ACT-R) is a widely used cognitive architecture that defines the cognitive and perceptual operations of the human mind. This study combines the ACT-R and EEG-based cortex-level connectivity to highlight the relationship between ACT-R modules during the EEG-based n-back task (for validating working memory performance). Initially, the source localization method is performed on the EEG signal, and the mapping between ACT-R modules and corresponding brain scouts (on the cortex surface) is performed. Once the brain scouts are identified for ACT-R modules, then those scouts are called ACT-R scouts. The linear (Granger Causality: GC) and non-linear effective connectivity (Multivariate Transfer Entropy: MTE) methods are applied over the scouts’ time series data. From the GC and MTE analysis, for all n-back tasks, information flow is observed from the visual-to-imaginal ACT-R scout for storing the visual stimuli (i.e., input letter) in short-term memory. For 2 and 3-back tasks, causal flow exists from imaginal to retrieval ACT-R scout and vice-versa. Causal flow from procedural to the imaginal ACT-R scout is also observed for all workload levels to execute the set of productions. Identifying the relationship among ACT-R modules through scout-level connectivity in the cortical surface facilitates the effects of human cognition in terms of brain dynamics.

Intrinsic motivation in cognitive architecture: intellectual curiosity originated from pattern discovery

Intrinsic motivation in cognitive architecture: intellectual curiosity originated from pattern discovery

Personalized Model-Driven Interventions for Decisions From Experience.

Cognitive models that represent individuals provide many benefits for understanding the full range of human behavior. One way in which individual differences emerge is through differences in knowledge. In dynamic situations, where decisions are made from experience, models built upon a theory of experiential choice (instance-based learning theory; IBLT) can provide accurate predictions of individual human learning and adaptivity to changing environments. Here, we demonstrate how an instance-based learning (IBL) cognitive model, implemented in a cognitive architecture (Adaptive Control of Thought-Rational), can be used to model an individual's decisions in a cybersecurity defense task, accounting for both population average and individual variances. The same IBL model structure with identical architectural parameters generates the full range of human behavior through stochastic memory retrieval processes operating over and contributing to unique experiences. Recurrence quantification analyses allow us to look beyond average behavior between and within individuals to sequential patterns of trial-to-trial behavior. We show how model-tracing and knowledge-tracing techniques can be used to align the model to an individual in real time to drive adaptive and personalized signaling algorithms for a cybersecurity defense system. We also present a method for introspecting into the cognitive model to gain further insight into the cognitive salience of features factored into individual decisions. The combination of techniques provides a blueprint for personalized modeling of individuals. We discuss the results and implications of this adaptive and personalized method for cybersecurity defense and more generally for intelligent artifacts tailored to individual differences in domains such as human-machine teaming.

Predicting and Understanding Human Action Decisions: Insights from Large Language Models and Cognitive Instance-Based Learning

Large Language Models (LLMs) excel in tasks from translation to complex reasoning. For AI systems to help effectively, understanding and predicting human behavior and biases is essential. However, it remains an open question whether LLMs can achieve this goal. This paper addresses this gap by leveraging the reasoning and generative capabilities of LLMs to predict human behavior in two sequential decision-making tasks. These tasks involve balancing between exploratory and exploitative actions and handling delayed feedback, which is essential for simulating real-life decision processes. We compare the performance of LLMs with a cognitive instance-based learning (IBL) model, which imitates human experiential decision-making. Our findings indicate that LLMs excel at rapidly incorporating feedback to enhance prediction accuracy. In contrast, the IBL model better accounts for human exploratory behaviors and effectively captures loss aversion bias — the tendency to choose a sub-optimal goal with fewer step-cost penalties rather than exploring to find the optimal choice, even with limited experience. The results highlight the benefits of integrating LLMs with cognitive architectures, suggesting that this synergy could enhance the modeling and understanding of complex human decision-making patterns.

Hybrid lane change strategy of autonomous vehicles based on SOAR cognitive architecture and deep reinforcement learning

Research on lane change strategies for autonomous vehicles holds paramount importance in optimizing traffic flow efficiency, enhancing driving safety, and adapting to complex traffic environments. While numerous rule-based or machine-learning approaches have been explored to tackle the challenge of lane change on highways, they frequently exhibit limited performance owing to the complexity of driving environments. This study proposes a novel lane change strategy for autonomous vehicles, which utilizes a hybrid framework integrating the SOAR cognitive architecture and deep reinforcement learning (DRL) to address the lane change challenge on highways. First, we introduce a rule extraction algorithm, the RuleCOSI+, which is based on tree ensemble algorithms, designed to extract concise lane change rules from large-scale human driving data. These straightforward rules, together with traffic regulations and safety rules, constitute the long-term memory of the SOAR cognitive architecture, enabling transparent decision-making processes. Next, by analyzing the clipping mechanism of the proximal policy optimization (PPO) algorithm, we propose an Adaptive Clipping PPO (ACPPO) algorithm which is based on the importance of samples. This algorithm adopts different clipping strategies for SOAR samples and ACPPO samples during the training process, enabling the algorithm to more effectively utilize samples with different levels of importance. Then, we propose a hybrid decision-making algorithm: SOAR-ACPPO, which combines the SOAR cognitive architecture with the ACPPO algorithm. This algorithm leverages SOAR’s prior knowledge to effectively and safely guide agent learning. Finally, by selecting appropriate intervention probability and weaning strategy, the system avoids inappropriate knowledge intervention and ensures adequate environment exploration. Simulation experiments conducted using the CARLA simulator illustrate that the proposed strategy not only improves model learning efficiency but also enhances driving efficiency and safety. Additionally, it demonstrates a certain degree of human-like characteristics and interpretability.

To the problem of digital immortality

The idea of reproducing personality by means of digital (neural network) technologies (“digital immortality”), together with the concept of Digital Twin (DT), still attracts great attention. Recent advances in the DT industry permit to expect the production of perfect “mirror” DT in the near future. We argue that the “immortality in the memory of other people” could be approached quite closely due to creating an analogue of personal DT by simulating the personality. For this purpose, it is necessary to compose the “constructive portrait” of a chosen person (by extracting the key features and traits of personality) and try to reproduce it by means of a chosen model. We are developing an original model Natural Constructive Cognitive Architecture (NCCA) that inherently provides the interpretation of logical and intuitive thinking, subconscious, etc. This model should be adjusted to specific set of knowledge inherent in a particular person (books, films, photographs, etc.), with an emphasis on personal lexicons (verbal, emotion, behavioral). NCCA contains a large set of free model parameters, which enables us to reproduce a wide range of personality features, from thinking style to temperament. It is shown that popular Generative Pre-trained Transformers (GPTs) have much in common with NCCA and could be adapted and used as an analog of DT of a specific person. We argue that the proposed program would provide the possibility to create an analog of DT, which could give an impression (at least, an illusion) of communication with the desired specific person.

Artificial General Intelligence (AGI): A Comprehensive Review

Artificial General Intelligence (AGI) represents a significant leap in the field of artificial intelligence, defined by its ability to perform any intellectual task that a human can. Unlike narrow AI, which is task-specific, AGI is characterized by versatility, adaptability, autonomy, and reasoning capabilities. This comprehensive review explores the defining features of AGI, including its cognitive flexibility and capacity for autonomous decision-making and problem-solving. The current state of AGI research is examined, highlighting key areas such as cognitive architectures, neuroscience-inspired AI, and machine learning and deep learning advancements. The review also addresses the technical challenges and ethical considerations associated with AGI development, including potential impacts on employment, privacy, and security, as well as the necessity for robust safety and control measures. Looking ahead, the potential benefits of AGI in various domains, such as healthcare, climate change, education, and economic growth, are discussed. Finally, the importance of establishing ethical frameworks and governance structures to guide AGI development and usage is emphasized. By addressing these challenges and fostering collaboration among scientists, policymakers, and society, AGI can be developed and deployed to benefit humanity as a whole.

The effect of chess on cognition: a graph theory study on cognitive data.

We aimed to advance our understanding of the effect of chess on cognition by expanding previous univariate studies with the use of graph theory on cognitive data. Specifically, we investigated the cognitive connectome of adult chess players. We included 19 chess players and 19 controls with ages between 39 and 69 years. Univariate analysis and graph theory included 27 cognitive measures representing multiple cognitive domains and subdomains. Graph analysis included global and nodal measures of integration, segregation, and centrality. We also performed an analysis of community structures to gain an additional understanding of the cognitive architecture of chess players. The analysis of global graph measures showed that chess players had a higher local efficiency than controls at the cost of a lower global efficiency, which did not permeate segregation aspects of their connectome. The nodal graph measures showed that executive/attention/processing speed and visuoconstructive nodes had a central role in the connectome of chess players. The analysis of communities showed that chess players had a slightly reorganized cognitive architecture into three modules. These graph theory findings were in the context of better cognitive performance in chess players than controls in visuospatial abilities. We conclude that the cognitive architecture of chess players is slightly reorganized into functionally and anatomically coherent modules reflecting a distinction between visual, verbal, and executive/attention/processing speed-related functions, perhaps reminiscent of right hemisphere and left hemisphere subnetworks orchestrated by the frontal lobe and its white matter connections.

Synthesizing the temporal self: robotic models of episodic and autobiographical memory.

Episodic memories are experienced as belonging to a self that persists in time. We review evidence concerning the nature of human episodic memory and of the sense of self and how these emerge during development, proposing that the younger child experiences a persistent self that supports a subjective experience of remembering. We then explore recent research in cognitive architectures for robotics that has investigated the possibility of forms of synthetic episodic and autobiographical memory. We show that recent advances in generative modeling can support an understanding of the emergence of self and of episodic memory, and that cognitive architectures which include a language capacity are showing progress towards the construction of a narrative self with autobiographical memory capabilities for robots. We conclude by considering the prospects for a more complete model of mental time travel in robotics and the implications of this modeling work for understanding human episodic memory and the self in time. This article is part of the theme issue 'Elements of episodic memory: lessons from 40 years of research'.

Explaining Crisis Situations via a Cognitive Model of Attention

Decision making in critical situations is a complex process. There are many processes to consider. This paper describes a theoretical approach to researching attentional processes and automatic unconscious processes in terms of metacognition. An application of the approach is presented to explain decision making and metacognition as a solution for ineffective cognitive biases during a crisis situation. Evidence is presented from studies on neuropsychology, cognitive control, and cognitive architectures. An application of the recently formulated semiotic methodology is implemented that allows the design of conceptual models of Attention as Action. The formulation of a general model of attentional processes is based on a set of rules. The crisis phenomenon, as the crisis situation trigger, is semiotically described and applied as insight for a crisis information system design that prompts its users toward self-aware internal decision making. The research conducted evidently shows how the approach can explain the design of several cognitive architectures. Pointing toward metacognition as a solution to a crisis phenomenon and cognitive biases, the paper shows that understanding human cognitive and behavioral processes can significantly improve management in a critical infrastructure crisis situation.

Guessing Human Intentions to Avoid Dangerous Situations in Caregiving Robots

The integration of robots into social environments necessitates their ability to interpret human intentions and anticipate potential outcomes accurately. This capability is particularly crucial for social robots designed for human care, as they may encounter situations that pose significant risks to individuals, such as undetected obstacles in their path. These hazards must be identified and mitigated promptly to ensure human safety. This paper delves into the artificial theory of mind (ATM) approach to inferring and interpreting human intentions within human–robot interaction. We propose a novel algorithm that detects potentially hazardous situations for humans and selects appropriate robotic actions to eliminate these dangers in real time. Our methodology employs a simulation-based approach to ATM, incorporating a “like-me” policy to assign intentions and actions to human subjects. This strategy enables the robot to detect risks and act with a high success rate, even under time-constrained circumstances. The algorithm was seamlessly integrated into an existing robotics cognitive architecture, enhancing its social interaction and risk mitigation capabilities. To evaluate the robustness, precision, and real-time responsiveness of our implementation, we conducted a series of three experiments: (i) A fully simulated scenario to assess the algorithm’s performance in a controlled environment; (ii) A human-in-the-loop hybrid configuration to test the system’s adaptability to real-time human input; and (iii) A real-world scenario to validate the algorithm’s effectiveness in practical applications. These experiments provided comprehensive insights into the algorithm’s performance across various conditions, demonstrating its potential for improving the safety and efficacy of social robots in human care settings. Our findings contribute to the growing research on social robotics and artificial intelligence, offering a promising approach to enhancing human–robot interaction in potentially hazardous environments. Future work may explore the scalability of this algorithm to more complex scenarios and its integration with other advanced robotic systems.

Social–Emotional Conversational Agents Based on Cognitive Architectures and Machine Learning

Social–Emotional Conversational Agents Based on Cognitive Architectures and Machine Learning

The Cognitive Architecture of Infant Attachment.

Meta-analytic evidence indicates that the quality of the attachment relationship that infants establish with their primary caregiver has enduring significance for socioemotional and cognitive outcomes. However, the mechanisms by which early attachment experiences contribute to subsequent development remain underspecified. According to attachment theory, early attachment experiences become embodied in the form of cognitive-affective representations, referred to as internal working models (IWMs), that guide future behavior. Little is known, however, about the cognitive architecture of IWMs in infancy. In this article, we discuss significant advances made in the field of infant cognitive development and propose that leveraging insights from this research has the potential to fundamentally shape our understanding of the cognitive architecture of attachment representations in infancy. We also propose that the integration of attachment research into cognitive research can shed light on the role of early experiences, individual differences, and stability and change in infant cognition, as well as open new routes of investigation in cognitive studies, which will further our understanding of human knowledge. We provide recommendations for future research throughout the article and conclude by using our collaborative research as an example.

Brain-consistent architecture for imagination.

Imagination represents a pivotal capability of human intelligence. To develop human-like artificial intelligence, uncovering the computational architecture pertinent to imaginative capabilities through reverse engineering the brain's computational functions is essential. The existing Structure-Constrained Interface Decomposition (SCID) method, leverages the anatomical structure of the brain to extract computational architecture. However, its efficacy is limited to narrow brain regions, making it unsuitable for realizing the function of imagination, which involves diverse brain areas such as the neocortex, basal ganglia, thalamus, and hippocampus. In this study, we proposed the Function-Oriented SCID method, an advancement over the existing SCID method, comprising four steps designed for reverse engineering broader brain areas. This method was applied to the brain's imaginative capabilities to design a hypothetical computational architecture. The implementation began with defining the human imaginative ability that we aspire to simulate. Subsequently, six critical requirements necessary for actualizing the defined imagination were identified. Constraints were established considering the unique representational capacity and the singularity of the neocortex's modes, a distributed memory structure responsible for executing imaginative functions. In line with these constraints, we developed five distinct functions to fulfill the requirements. We allocated specific components for each function, followed by an architectural proposal aligning each component with a corresponding brain organ. In the proposed architecture, the distributed memory component, associated with the neocortex, realizes the representation and execution function; the imaginary zone maker component, associated with the claustrum, accomplishes the dynamic-zone partitioning function; the routing conductor component, linked with the complex of thalamus and basal ganglia, performs the manipulation function; the mode memory component, related to the specific agranular neocortical area executes the mode maintenance function; and the recorder component, affiliated with the hippocampal formation, handles the history management function. Thus, we have provided a fundamental cognitive architecture of the brain that comprehensively covers the brain's imaginative capacities.

A Workflow for Building Computationally Rational Models of Human Behavior

Computational rationality explains human behavior as arising due to the maximization of expected utility under the constraints imposed by the environment and limited cognitive resources. This simple assumption, when instantiated via partially observable Markov decision processes (POMDPs), gives rise to a powerful approach for modeling human adaptive behavior, within which a variety of internal models of cognition can be embedded. In particular, such an instantiation enables the use of methods from reinforcement learning (RL) to approximate the optimal policy solution to the sequential decision-making problems posed to the cognitive system in any given setting; this stands in contrast to requiring ad hoc hand-crafted rules for capturing adaptive behavior in more traditional cognitive architectures. However, despite their successes and promise for modeling human adaptive behavior across everyday tasks, computationally rational models that use RL are not easy to build. Being a hybrid of theoretical cognitive models and machine learning (ML) necessitates that model building take into account appropriate practices from both cognitive science and ML. The design of psychological assumptions and machine learning decisions concerning reward specification, policy optimization, parameter inference, and model selection are all tangled processes rife with pitfalls that can hinder the development of valid and effective models. Drawing from a decade of work on this approach, a workflow is outlined for tackling this challenge and is accompanied by a detailed discussion of the pros and cons at key decision points.

A universal knowledge model and cognitive architectures for prototyping AGI

The article identified 56 cognitive architectures for creating general artificial intelligence (AGI) and proposed a set of interrelated functional blocks that an agent approaching AGI in its capabilities should possess. Since the required set of blocks is not found in any of the existing architectures, the article proposes a reference cognitive architecture for intelligent systems approaching AGI in their capabilities. As one of the key solutions within the framework of the architecture, a universal method of knowledge representation is proposed, which allows combining various non-formalized, partially and fully formalized methods of knowledge representation in a single knowledge base, such as texts in natural languages, images, audio and video recordings, graphs, algorithms, databases, neural networks, knowledge graphs, ontologies, frames, essence-property-relation models, production systems, predicate calculus models, conceptual models, and others. To combine and structure various fragments of knowledge, archigraph model are used, constructed as a development of annotated metagraphs. As other components, the reference cognitive architecture being developed includes following modules: machine consciousness, machine subconsciousness, interaction with the external environment, a goal management, an emotional control, social interaction, reflection, ethics, worldview, learning, monitoring, statement problems, solving problems, self-organization and meta learning. Based on the composition of the proposed reference architecture modules, existing cognitive architectures containing the following modules were analyzed: machine consciousness, machine subconsciousness, reflection, worldview.

Contrast Information Dynamics: A Novel Information Measure for Cognitive Modelling.

We present contrast information, a novel application of some specific cases of relative entropy, designed to be useful for the cognitive modelling of the sequential perception of continuous signals. We explain the relevance of entropy in the cognitive modelling of sequential phenomena such as music and language. Then, as a first step to demonstrating the utility of constrast information for this purpose, we empirically show that its discrete case correlates well with existing successful cognitive models in the literature. We explain some interesting properties of constrast information. Finally, we propose future work toward a cognitive architecture that uses it.

Towards mechanistic investigations of numerical and music cognition.

Are there cognitive connections between humans' ability to make music and their understanding of math and numbers? This question has motivated centuries of speculation across the fields of philosophy and education and an increasing number of empirical studies of the topic. We review research at the intersection of numerical and music cognition, and establish its relevance both to the applied sphere (e.g., education) and to core theoretical issues in the cognitive sciences. Next, we identify notable limitations within the literature, and attempt to explain some of the likely causes of-and solutions to-these limitations. Finally, we propose specific themes of focus (spatialization and the mapping between symbolic and nonsymbolic representations) for future research aimed at understanding whether there is shared cognitive architecture for reasoning about number and math. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

Social perception of animacy: Preferential attentional orienting to animals links with autistic traits

Animate cues enjoy priority in attentional processes as they carry survival-relevant information and herald social interaction. Whether and in what way such an attention effect is associated with more general aspects of social cognition remains largely unexplored. Here we investigated whether the attentional preference for animals varies with observers' autistic traits — an indicator of autism-like characteristics in general populations related to one's social cognitive abilities. Using the dot-probe paradigm, we found that animal cues can rapidly and persistently recruit preferential attention over inanimate ones in observers with relatively low, but not high, autistic traits, as measured by Autism-Spectrum Quotient (AQ). Moreover, individual AQ scores were negatively correlated with the attentional bias toward animals, especially at the early orienting stage. These results were not simply due to low-level visual factors, as inverted or phase-scrambled pictures did not yield a similar pattern. Our findings demonstrate an automatic and enduring attentional bias beneficial to both rapid detection and continuous monitoring of animals and reveal its link with autistic traits, highlighting the critical role of animacy perception in the architecture of social cognition.

Temporal heterogeneity in cognitive architectures

In 2020, Mc Fadden published an article in which he discusses how algorithms can be encoded in time and space. By analyzing the topology of the cytoarchitecture of the brain, cognitive architectures can understand the underlying mechanisms that have led to the development of human intelligence in space. In this study, our focus lies in investigating temporal heterogeneity as a mechanism that the brain could have developed not solely as a biological constraint, but also as an evolutionary advantage. To accomplish this, we employed virtual agents within a virtual environment and constructed a prototype cognitive architecture. Subsequently, we compared the benefits and drawbacks of having this cognitive architecture operate under a model of temporal heterogeneity versus one characterized by temporal homogeneity. At the conclusion of the article, we present the results obtained from two perspectives. From a quantitative standpoint, we contrast the agents’ adaptation to the environment based on the cognitive architecture model employed by each agent. On this front, we found evidence that temporal heterogeneity might be useful in finding parameter optimizations faster, amongst other benefits. From a qualitative perspective, we examine the potential of this model to explore the cognitive processes of the virtual agents, concluding that a different representation of percepts is needed, which we further discuss.