Search

This literature review explores the role of attractor neural networks (ANNs) in modeling psychological processes in artificial and biological systems. By synthesizing research from dynamical systems theory, psychology, and computational neuroscience, the review provides an overview of the current understanding of ANN function in memory formation, memory reinforcement, retrieval, and forgetting. Key mathematical foundations of ANNs, including dynamical systems theory and energy functions, are discussed to explain the behavior and stability of these networks. The review also examines empirical applications of ANNs in cognitive processes such as semantic memory and episodic recall, as well as highlighting the hippocampus' role in pattern separation and completion. The review addresses challenges like catastrophic forgetting and noise effects on memory retrieval. By identifying gaps between theoretical models and empirical findings, it highlights the interdisciplinary nature of ANN research and suggests future areas for exploration.

Pyrone-2,4-dicarboxylic acid (PDC) is a valuable polymer precursor that can be derived from the microbial degradation of lignin. The key enzyme in the microbial production of PDC is 4-carboxy-2-hydroxymuconate-6-semialdehyde (CHMS) dehydrogenase, which acts on the substrate CHMS. We present the crystal structure of CHMS dehydrogenase (PmdC from Comamonas testosteroni) bound to the cofactor NADP, shedding light on its three-dimensional architecture, and revealing residues responsible for binding NADP. Using a combination of structural homology, molecular docking, and quantum chemistry calculations, we have predicted the binding site of CHMS. Key histidine residues in a conserved sequence are identified as crucial for binding the hydroxyl group of CHMS and facilitating dehydrogenation with NADP. Mutating these histidine residues results in a loss of enzyme activity, leading to a proposed model for the enzyme's mechanism. These findings are expected to help guide efforts in protein and metabolic engineering to enhance PDC yields in biological routes to polymer feedstock synthesis.

In many organisms, most notably Drosophila, homologous chromosomes associate in somatic cells, a phenomenon known as somatic pairing, which takes place without double strand breaks or strand invasion, thus requiring some other mechanism for homologs to recognize each other. Several studies have suggested a "specific button" model, in which a series of distinct regions in the genome, known as buttons, can associate with each other, mediated by different proteins that bind to these different regions. Here, we use computational modeling to evaluate an alternative "button barcode" model, in which there is only one type of recognition site or adhesion button, present in many copies in the genome, each of which can associate with any of the others with equal affinity. In this model, buttons are nonuniformly distributed, such that alignment of a chromosome with its correct homolog, compared with a nonhomolog, is energetically favored; since to achieve nonhomologous alignment, chromosomes would be required to mechanically deform in order to bring their buttons into mutual register. By simulating randomly generated nonuniform button distributions, many highly effective button barcodes can be easily found, some of which achieve virtually perfect pairing fidelity. This model is consistent with existing literature on the effect of translocations of different sizes on homolog pairing. We conclude that a button barcode model can attain highly specific homolog recognition, comparable to that seen in actual cells undergoing somatic homolog pairing, without the need for specific interactions. This model may have implications for how meiotic pairing is achieved.

Smoke taint in wine has become a critical issue in the wine industry due to its significant negative impact on wine quality. Data-driven approaches including univariate analysis and predictive modeling are applied to a data set containing concentrations of 20 VOCs in 48 grape samples and 56 corresponding wine samples with a taster-evaluated smoke taint index. The resulting models for predicting the smoke taint index of wines are highly predictive when using as inputs VOC concentrations after log conversion in both grapes and wines (Pearson Correlation Coefficient PCC = 0.82; R2 = 0.68) and less so when only grape VOCs are used (Pearson Correlation Coefficient PCC = 0.76; R2 = 0.56), and the classification models also show the capacity for detecting smoke-tainted wines using both wine and grape VOC concentrations (Recall = 0.76; Precision = 0.92; F1 = 0.82) or using only grape VOC concentrations (Recall = 0.74; Precision = 0.92; F1 = 0.80). The performance of the predictive model shows the possibility of predicting the smoke taint index of the wine and grape samples before fermentation. The corresponding code of data analysis and predictive modeling of smoke taint in wine is available in the Github repository (https://github.com/IBPA/smoke_taint_prediction).

Earthquakes present worldwide risk to economic and human safety. The 2023 earthquakes in Turkiye provided a reminder of the potential for catastrophic consequences with 50,700 deaths and 15.7 million people affected. The ability to predict ground motions and infrastructure damage for earthquakes continues to be a challenging problem for scientists and engineers. Until now, estimates of ground motions have been performed empirically by looking at sparse data from past earthquakes. This approach can provide statistical information on intensity amplitudes but cannot inform site-specific ground motions essential to developing the most effective resilience. Interest has grown in large-scale computational models to simulate earthquakes at regional scale. The U.S. Department of Energy EarthQuake SIMulation (EQSIM) framework was developed for regional-scale earthquake simulations at unprecedented fidelity, taking advantage of emerging GPU-accelerated systems. This article describes the EQSIM workflow and demonstrates regional-scale simulations with the new computational capability available to scientists in their quest to mitigate future disasters.

Models have historically represented fractured porous media with continuum descriptions that characterize the media using bulk parameters. The impact of small-scale features is not captured in these models, although they may be controlling the performance of subsurface applications. Pore-scale models can simulate processes in small-scale features by representing the pore space geometry explicitly but are computationally expensive for large domains. The alternative multiscale approach entails the combination of pore-scale and continuum-scale descriptions in a single framework. We use Chombo-Crunch, a computational capability that discretizes complex geometries with an adaptive embedded boundary method to contrast these two approaches. Chombo-Crunch takes advantage of recent computational performance and memory bandwidth improvements resulting from the emergence of exascale computing resources. These combined improvements enable the efficient simulation of reactive transport in fractured media with a high degree of fidelity and the ability to capture the control small-scale processes exert on the overall medium evolution.

Anxiety disorders affect millions of people worldwide and present a challenge in neuroscience research because of their substantial heterogeneity in clinical presentation. While a great deal of progress has been made in understanding the neurobiology of fear and anxiety, these insights have not led to effective treatments. Understanding the relationship between phenotypic heterogeneity and the underlying biology is a critical first step in solving this problem. We show translation, reverse translation, and computational modeling can contribute to a refined, cross-species understanding of fear and anxiety as well as anxiety disorders. More specifically, we outline how animal models can be leveraged to develop testable hypotheses in humans by using targeted, cross-species approaches and ethologically informed behavioral paradigms. We discuss reverse translational approaches that can guide and prioritize animal research in nontraditional research species. Finally, we advocate for the use of computational models to harmonize cross-species and cross-methodology research into anxiety. Together, this translational neuroscience approach will help to bridge the widening gap between how we currently conceptualize and diagnose anxiety disorders, as well as aid in the discovery of better treatments for these conditions.

We present motivation and practical steps necessary to find parameter estimates of joint models of behavior and neural electrophysiological data. This tutorial is written for researchers wishing to build joint models of human behavior and scalp and intracranial electroencephalographic (EEG) or magnetoencephalographic (MEG) data, and more specifically those researchers who seek to understand human cognition. Although these techniques could easily be applied to animal models, the focus of this tutorial is on human participants. Joint modeling of M/EEG and behavior requires some knowledge of existing computational and cognitive theories, M/EEG artifact correction, M/EEG analysis techniques, cognitive modeling, and programming for statistical modeling implementation. This paper seeks to give an introduction to these techniques as they apply to estimating parameters from neurocognitive models of M/EEG and human behavior, and to evaluate model results and compare models. Due to our research and knowledge on the subject matter, our examples in this paper will focus on testing specific hypotheses in human decision-making theory. However, most of the motivation and discussion of this paper applies across many modeling procedures and applications. We provide Python (and linked R) code examples in the tutorial and appendix. Readers are encouraged to try the exercises at the end of the document.

Exascale supercomputers are first-of-their-kind instruments with potentially paradigm-shifting capabilities. However, enabling complex applications at scale and high performance can be difficult, requiring hard work and deep technical understanding. Specific programming models are often needed to achieve the desired performance, and, early on, systems are undergoing debugging and rapid development. This makes it challenging to run on new systems quickly following their deployment. The Application Integration (AppInt) area of the Exascale Computing Project was designed to be the integration point between applications, supporting software, system environments, high-performance computing facilities, and vendors. We describe how the AppInt team addressed these challenges while also promoting the use of portable and sustainable programming models and helping harden systems prior to general availability.

The Exascale Computing Project software deployment effort developed and advanced DevOps capabilities. One goal was to enable robust continuous integration (CI) workflows that span the protected high-performance computing environments found within many of the U.S. Department of Energy's (DOE's) national laboratories. This article highlights several challenges encountered with enabling automation, such as charging models for CI jobs and meeting individualized security requirements that revolve around strongly associating running code with a human identity. It also describes how the Jacamar CI tool evolved to meet later requirements and became a key aspect of the solutions currently offered. Derived from this experience, we offer a conceptual framework for understanding current and future CI challenges at DOE facilities and offer suggestions for long-term solutions.

Recent work in reinforcement learning has leveraged symmetries in the model to improve sample efficiency in training a policy. A commonly used simplifying assumption is that the dynamics and reward both exhibit the same symmetry; however, in many real-world environments, the dynamical model exhibits symmetry independent of the reward model. In this letter, we assume only the dynamics exhibit symmetry, extending the scope of problems in reinforcement learning and learning in control theory to which symmetry techniques can be applied. We use Cartan's moving frame method to introduce a technique for learning dynamics that, by construction, exhibit specified symmetries. Numerical experiments demonstrate that the proposed method learns a more accurate dynamical model.

Alfred Hitchcock, film director and "Master of Suspense," observed that terror is not driven by a negative event, but "only in the anticipation of it." This observation is not restricted to the movies: Anxiety builds when we anticipate uncertain negative events, and heightened reactivity during uncertain threat anticipation is a transdiagnostic marker of anxiety (Grupe & Nitschke, 2013; Holley & Fox, 2022; Hur et al., 2020; Krain et al., 2008; Simmons et al., 2008; Yassa et al., 2012). Here, we manipulate the temporal dynamics of an uncertain threat to demonstrate how the evolving expectation of threat can lead people to forgo rewards and experience fear/anxiety. Specifically, we show that increased "hazard rate," which can build during periods of uncertainty, promotes a tendency to avoid threatening contexts while increasing fear/anxiety. These results provide insight into why the anticipation of temporally uncertain threats elicits fear/anxiety, and reframe the underlying causes of related psychopathology.

Ion channels are critical drug targets for a range of pathologies, such as epilepsy, pain, itch, autoimmunity, and cardiac arrhythmias. To develop effective and safe therapeutics, it is necessary to design small molecules with high potency and selectivity for specific ion channel subtypes. There has been increasing implementation of structure-guided drug design for the development of small molecules targeting ion channels. We evaluated the performance of two RosettaLigand docking methods, RosettaLigand and GALigandDock, on the structures of known ligand-cation channel complexes. Ligands were docked to voltage-gated sodium (NaV), voltage-gated calcium (CaV), and transient receptor potential vanilloid (TRPV) channel families. For each test case, RosettaLigand and GALigandDock methods frequently sampled a ligand-binding pose within a root mean square deviation (RMSD) of 1-2 Å relative to the experimental ligand coordinates. However, RosettaLigand and GALigandDock scoring functions cannot consistently identify experimental ligand coordinates as top-scoring models. Our study reveals that the proper scoring criteria for RosettaLigand and GALigandDock modeling of ligand-ion channel complexes should be assessed on a case-by-case basis using sufficient ligand and receptor interface sampling, knowledge about state-specific interactions of the ion channel, and inherent receptor site flexibility that could influence ligand binding.

To conceptualise biological and mental processes, often a dynamical systems perspective is suggested. In addition to dynamics, the structure of the contextual makeup or world configuration (of an organism or brain) plays a crucial role too, as well as adaptivity of the processes. This paper provides a conceptual perspective where the structure, dynamics, and adaptivity of these processes are distinguished and related to each other via adaptive dynamical systems. Moreover, it is shown how networks can be used to represent this conceptual perspective. Here an adaptive dynamical system of any order of adaptivity can be covered where any level can exert control over the level below. The approach is illustrated by case studies for higher-order adaptive evolutionary processes. One of these case studies shows a fifth-order adaptive dynamical system that models how due to bad environmental influences at a young age, epigenetic effects can lead to a lifelong mental disorder.

In-context learning refers to the ability of a neural network to learn from information presented in its context. While traditional learning in neural networks requires adjusting network weights for every new task, in-context learning operates purely by updating internal activations without needing any updates to network weights. The emergence of this ability in large language models has led to a paradigm shift in machine learning and has forced researchers to reconceptualize how they think about learning in neural networks. Looking beyond language models, we can find in-context learning in many computational models relevant to cognitive science, including those that emerge from meta-learning.

In recent years, the field of computer graphics has achieved its longstanding dream of photorealism: modern graphics algorithms produce images that are indistinguishable from reality. Much like art at the advent of photography, then, computer graphics is now turning its gaze to the beholder: researchers are increasingly looking to cognitive science to engineer new modes of visual expression. Recent work has sought to apply insights from cognitive science to a variety of traditional graphics topics: from taking a perceptual approach to perspective, to studying the theory of mind behind animation, to applying theories of abstraction learning to build tools for geometry processing. At the same time, a wave of recent work in cognitive science has addressed fundamental questions about visual expression: for example, how humans understand and create sketches, shapes, and symbols. The field has also benefited greatly from tools and methods from computer graphics: differentiable rendering, physics simulation, and game engines have become indispensable in modeling human perception and intuitive physics. Recognizing this growing interdisciplinary exchange of ideas, we are proposing a workshop to begin building formal bridges between the cognitive science and computer graphics communities.

The field of psychometrics has undergone substantial evolution over the past several decades, both in terms of advances in methodology and improved software and hardware for deploying new methods. Despite these strides, many of these developments have not been integrated into the broader field of psychology, as highlighted by Embretson (2005) and Borsboom (2006). Understanding and incorporating these psychometric advances is crucial to enable cognitive scientists to address growing concerns about validity and reliability, as well as to develop robust theoretical frameworks for understanding cognition.

We have a simple thesis: the relationship between academic and industry-based cognitive science is broken, but can be fixed. Over the last few decades, there has been a huge increase in the representation of cognitive science in industry. Beyond just machine learning, businesses are increasingly interested in human behavior and cognitive processes. Large proportions of our Ph.D. students, post-docs, and even faculty choose to go through a largely one-way door to corporate jobs in data science, behavioral experimentation, machine learning, user experience, and elsewhere. Currently, people who choose industry careers often lose their social and intellectual networks and their ability to return to tenure-track positions. Valuable insights from industry about memory, decision-making, learning, emotion, distributed cognition, and much more never return to the academic community. We believe that deep, theory driven, theory building work is being done in industry settings–and that the rift between communities makes all our work less effective

Predicting the actions and reactions of others is crucial to suc- cessful social interaction. When deciding whether to bluff in a game of poker, we consider the chances that the other players will fold or continue to play and unmask our bluff. When deciding whether to tell our boss that their plans are likely to have adverse effects, we consider a range of reac- tions, from being grateful for our honesty to being dismissed out of spite. Such predictions are highly uncertain and com- plex, not least because the other's (re)actions usually result from them making equally complex and uncertain inferences about us. Nevertheless, we are often remarkably successful – although sometimes utterly wrong – in our social inferences. How do we explain these successes and failures?

Humans engage in a wide variety of search behaviors during their lifetime. These search behaviors may be external such as foraging for food in the wild, or searching for mates, or internal, such as searching for concepts in memory. Decades of work on search processes among humans has suggested that these two types of search behaviors share common characteristics and physiological mechanisms. Despite this progress, understanding the dynamics of internal search is an ongoing challenge in the field.

Cognitive processes underlie economic relations. In this paper, we develop a conceptual, mathematical, and computational framework for modeling market exchange as a series of dynamically interacting cognitive processes. Specifically, we show how advertisers can build trust and gain confidence in their pricing power to the point that they erode trust and undermine the efficacy of their advertising. Customers conversely orient towards advertisers seeking information or turn away from them as unreliable communicators. These behaviors and the patterns they generate occur inside a state space of unallocated perceived value. They constitute a small subset of the full range of possible strategic and adaptive responses that define cognitive microeconomics.

Adults are known to have superior working memory to children, but whether this improvement is driven primarily by differences in storage capacity or attentional control is debated. In particular, the understanding of how capacity and control influence the development of working memory is hampered by the fact that most theorizing about the effect of variation in either on behavior has been verbal. To address this, we extended a computational model of working memory to clearly separate the contributions of capacity and control, fitting the model to a recent developmental study. We find that the combined influence of capacity and control on working memory may be more complicated than previously appreciated. In particular, the general pattern of qualitative differences between children and adults could be produced by increasing either capacity or control alone. These results point to a need for additional experimental paradigms to clearly parse the differential impact of working memory components.

Since their inception, drift diffusion models have been applied across a wide range of disciplines within psychology to uncover the mental processes that underlie perception, attention, and cognitive control. Our studies contribute to ongoing efforts to extend these models to abstract, social reasoning processes like moral or legal judgment. We presented participants with a set of social rules, while manipulating whether variousbehaviors violated the rule's letter and/or its purpose–––two independent standards by which to decide what constitutes a transgression. In this framework, cases that violate or comply with both a rule's text and its purpose can be seen as congruent or ‘easy' cases, and cases that elicit opposing verdicts as incongruent or ‘hard' cases–––in a manner analogous to widely-studied conflict tasks in cognitive psychology. We recorded 34,573 decisions made by 364 participants under soft time pressure, and investigated whether hierarchical drift diffusion modeling could explain various behavioral patterns in our data. This approach yielded three key insights: (1) judgments of conviction were faster than judgments of acquittal owing to an overall bias (z parameter) toward conviction; (2) incongruent cases produced longer reaction times than congruent cases (an interference effect), due to differences in the rate of evidence accumulation (v parameter) across case-types; and (3) increases in the ratio of congruent-to-incongruent cases amplified the interference effect on reaction times, by fostering greater response caution—revealed by a larger threshold (or a parameter). Thus, our studies document dissociable effects of the drift diffusion components on rule-based decision-making, and illustrate how the cognitive processes that subserve abstract and social decision-making tasks, such as the enforcement of communal and legal rules, may be illuminated through the drift diffusion framework.

Here we used deep neural network + hidden Markov model (DNN+HMM) to provide a computational account for the relationship among holistic processing (HP), face scanning pattern and face recognition performance. The model accounted for the positive associations between HP and eyes-focused face scanning pattern/face recognition performance observed in the literature regardless of the version of the composite task used to measure HP. Interestingly, we observed a quadratic relationship between HP and face scanning pattern, where models being highly eyes-focused or highly nose-focused had lower HP. By inspecting fixation locations and associated attention window size in the model and XAI methods, we found that the eyes- and nose-focused models both developed local and holistic internal representations during training, and their difference was in the temporal dynamics of how these representations were used. Our findings demonstrated how computational modeling could unravel the mechanisms underlying cognition not readily observable in human data.

Human learning does not stop at solving a single problem. Instead, we seek new challenges, define new goals, and come up with new ideas. What drives people to disrupt the existing conceptual landscape and create new things? Here, we examine the decision to create new things under different levels of expected returns. We formalize innovation as stochastically recombining existing ideas, where successful and more complex combinations generate higher returns. This formalization allows us to cast innovation-seeking as a Markov decision process, and derive optimal policies under different settings. Data collected through an online behavioral experiment confirm our prediction that people should invest more time and effort in seeking innovations when they know the chances of success are high and the potential new ideas would be rewarding. However, people also deviate from being optimal, both innovating more and less than they should in different settings.

Spontaneous recovery of fear after extinction is a well-established behavioral phenomenon. Different theories in psychology account for spontaneous recovery by proposing that it may result from temporal weighting, reduced processing of stimuli over time, enhanced salience of adverse events or return of the acquisition context. We propose a novel mechanism of spontaneous recovery: selective maintenance of adverse events, and ground this mechanism in a computational model of latent cause inference. To investigate the proposed mechanism, we collected behavioral data with an aversive conditioning and extinction task (N=280) and fit the data with computational models formalizing our and others' theories. Quantitative and qualitative model comparisons indicated that selective maintenance of adverse events accounts for spontaneous recovery better than alternative theories. As spontaneous recovery of fear after extinction can serve as a model of relapse after exposure therapy, we use this mechanistic understanding of spontaneous recovery to propose and simulate the effect of add-on interventions to prevent relapse after exposure therapy.

Comparison of visibility inferences across congenitally blind and sighted people provides insight into the contribution of first-person sensory experience to intuitive theories. We hypothesized that both groups understand others' visual experiences via an intuitive theory incorporating variables known to influence visual psychophysics (distance, looking duration, and feature size). Adults born blind (n=20) and sighted (n=40) listened to short scenarios that described an observer looking at another person from different distances and for varying durations. Participants rated how likely the observer would perceive appearance features of the person that varied in size (e.g., eye color vs. hat). A probabilistic formalization of intuitive visibility fit the ratings with high accuracy across scenarios and features. Model parameters were qualitatively identical across groups but blind adults weighted distance and size less. A quantitative and generative intuitive theory of vision develops without first-person sensory access, possibly through linguistic communication, and is fine-tuned by visual experience.

Previous research has shown that our subjective sense of certainty doesn't always accurately reflect the strength of the evidence that has been presented to us. We investigate several key factors that drive children's certainty using a Boolean concept learning task. We created an idealized learning model to predict children's accuracy and certainty during the experiment, given past evidence that they have seen in the task, and we compared its predictions with our behavioral results. Our results suggest that while predictors from the idealized learning model capture children's accuracy, behavioral predictors generated by the behavioral data can better predict children's certainty. We also show that younger children's certainty can be explained by the idealized learning model, while older children's certainty is primarily predicted by how well they observed themselves doing in the experiment.

Surprisal values from large language models (LLMs) have been used to model the amplitude of the N400.This ERP component is sensitive not only to contextual word expectancy but also to semantic association, such that unexpected but associated words do not always induce an N400 increase.While LLMs are also sensitive to association, it remains unclear how they behave in these cases.Moreover, another ERP component, the P600, has shown graded sensitivity to plausibility-driven expectancy, while remaining insensitive to association; however, its relationship to LLM surprisal is not well researched yet.In an rERP analysis, we evaluate surprisal values of two unidirectional transformers on their ability to model N400 and P600 effects observed in three German ERP studies isolating the effects of association, plausibility, and expectancy.We find that surprisal predicts an N400 increase for associated but implausible words, even when no such increase was observed in humans.Furthermore, LLM surprisal accounts for P600 effects elicited by violations of selectional restrictions, but captures neither P600 effects from more subtle script knowledge violations nor graded P600 modulations.The results of our investigation call into question the extent to which LLM surprisal offers an accurate characterisation of the functional generators of either the N400 or P600.

Computational psycholinguistics has traditionally employed a complexity metric called Node Count, which counts the number of syntactic nodes representing syntactic structures and predicts processing costs in human sentence processing. However, Node Count does not dissociate distinct syntactic operations deriving those syntactic structures, so that how much processing cost each syntactic operation induces remains to be investigated. In this paper, we introduce a novel complexity metric dubbed Composition Count, which counts the number of syntactic operations deriving syntactic structures, allowing us to understand the computational system of human sentence processing from the derivational, not representational, perspective. Specifically, employing Combinatory Categorial Grammar (CCG) which is equipped with multiple syntactic operations and thus suitable for the purpose here, we investigate (i) how much distinct syntactic operations of CCG contribute to predicting human reading times, and (ii) whether the same holds across languages. The results demonstrate that distinct syntactic operations of CCG have independent and cross-linguistic contributions to predicting human reading times, while Node Count turns out not to be robust cross-linguistically. In conclusion, these results strongly suggest the importance of Composition Count to dissociate distinct syntactic operations, not whole syntactic representations, and understand the computational system of human sentence processing.

Whereas cognitive models of learning often assume direct experience with both the features of an event and with a true label or outcome, much of everyday learning arises from hearing the opinions of others, without direct access to either the experience or the ground truth outcome. We consider how people can learn which opinions to trust in such scenarios by extending the hedge algorithm: a classic solution for learning from diverse information sources. We first introduce a semi-supervised variant we call the delusional hedge capable of learning from both supervised and unsupervised experiences. In two experiments, we examine the alignment between human judgments and predictions from the standard hedge, the delusional hedge, and a heuristic baseline model. Results indicate that humans effectively incorporate both labeled and unlabeled information in a manner consistent with the delusional hedge algorithm---suggesting that human learners not only gauge the accuracy of information sources but also their consistency with other reliable sources. The findings advance our understanding of human learning from diverse opinions, with implications for the development of algorithms that better capture how people learn to weigh conflicting information sources.

Children learn powerful internal models of the world around them from a few years of egocentric visual experience. Can such internal models be learned from a child's visual experience with highly generic learning algorithms or do they require strong inductive biases? Recent advances in collecting large-scale, longitudinal, developmentally realistic video datasets and generic self-supervised learning (SSL) algorithms are allowing us to begin to tackle this nature vs. nurture question. However, existing work typically focuses on image-based SSL algorithms and visual capabilities that can be learned from static images (e.g. object recognition), thus ignoring temporal aspects of the world. To close this gap, here we train self-supervised video models on longitudinal, egocentric headcam recordings collected from a child over a two year period in their early development (6-31 months). The resulting models are highly effective at facilitating the learning of action concepts from a small number of labeled examples; they have favorable data size scaling properties; and they display emergent video interpolation capabilities. Video models also learn more robust object representations than image-based models trained with the exact same data. These results suggest that important temporal aspects of a child's internal model of the world may be learnable from their visual experience using highly generic learning algorithms and without strong inductive biases.

During sentence processing, comprehenders have to maintain a mapping between lexical items and their position in the sentence (syntactic position). We propose a model of morpheme-position binding in working memory, based on models such as 'serial-order-in-a-box' and its SOB-complex-span version. Like those working memory models, our sentence processing version derives a range of attested memory interference effects from the process of item-position binding. We present simulation results capturing similarity-based interference and item-distortion. These two major classes of interference effects have not received a unified account before, and are not fully captured by cue-based retrieval models.

Decoding naturalistic stimuli from neural recordings is a significant challenge in systems neuroscience, primarily due to the high-dimensional and nonlinear nature of stimulus-response interactions, and is further exacerbated by the limited availability and noisiness of neural data. While contemporary approaches that incorporate generative models, such as Generative Adversarial Networks (GANs), attempt to address these issues by mapping neural responses to latent representations, they do not fully overcome these obstacles. In this work, we present a novel paradigm of differential neural decoding (dicoding) that focuses on the relative changes in response patterns, which not only expands the neural training data quadratically but also inherently denoises it. To determine the corresponding stimulus changes, this method leverages the Euclidean and feature-disentangled properties of the underlying latents through vector arithmetic. As such, we not only effectively exploit the latent space but also achieve semantically meaningful latent offsets in the context of the stimuli, resulting in improved sample efficiency. We trained a decoder to predict changes in latent vectors based on the corresponding changes in neural responses. The absolute latent vector itself was derived by vector addition of the predicted latent change (indicative of stimulus shift) to a reference latent, which was fed to the generator for the reconstruction of the perceived stimulus. Our results show that this geometrically principled approach facilitates more effective reconstruction of naturalistic stimuli from noisy and limited neural data.

We offer a computational framework for modeling explanation as cooperative rational communication. Under our framework, when an explainer is faced with a ``why?'' question, they reason about the question-asker's current mental model, and intervene on that mental model in order to maximize the listener's future utility. We instantiate our framework in a planning domain, and show that our framework can model human explanations about plans across a wide variety of scenarios.

Advancements in deep neural networks have led to significant progress in computer vision and natural language processing. These networks, trained on real-world stimuli, develop high-level feature representations of stimuli. It is hypothesized that these representations, stemming from different inputs, should converge into similar conceptual systems, as they reflect various perspectives of the same underlying reality. This paper examines the degree to which different conceptual systems can be aligned in an unsupervised manner, using feature-based representations from deep neural networks. Our investigation centers on the alignment between the image and word representations produced by diverse neural networks, emphasizing those trained via self-supervised learning methods. Subsequently, to probe comparable alignment patterns in human learning, we extend this examination to models trained on developmental headcam data from children. Our findings reveal a more pronounced alignment in models trained through self-supervised learning compared to supervised learning, effectively uncovering higher-level structural connections among categories. However, this alignment was notably absent in models trained with limited developmental headcam data, suggesting more data, more inductive biases, or more supervision are needed to establish alignment from realistic input.

Studies of meaning in human and primate communication face, in principle, similar methodological problems. In both cases, meaning is not observable directly, but must be inferred from more indirect sources, such as directly observable behavior. Recent work in probabilistic cognitive modeling of language use has therefore developed methods of inferring latent se- mantic meaning through the lens of a probabilistic model of language use. In this paper, we explore how to adapt such an approach for insightful investigations of primate communication. Towards this end, we develop a suitable probabilistic model of processes that generate communicative behavior by making use of functionally specified latent meaning representations. As a proof of concept, we apply this model to a rich, annotated data set of orangutan communicative dyadic interaction and conclude that explicit probabilistic modeling can provide additional insights for the study of animal communication pertaining to the context-dependent nature of signals and the gradual evolution of human communication systems.

Music-making relies on precise temporal control and mutual coordination among performers, particularly to maintain tempo. We evaluate the impact of human-machine interaction and rhythmic subdivisions on tempo change in musical trios. A synchronization-continuation task was performed by trios of human participants interacting with confederates or with algorithmic (i.e. machine) models. Sounded tone onsets were produced by a linear error-correction model, delay-coupled model, and Kuramoto model that replaced a human participant. Inter-onset intervals were examined from participants who performed rhythms in both in-phase and anti-phase conditions while a third group member was either a human or algorithmic model. Trios drifted toward faster tempi more when they contained a human than an algorithmic model. Tempo drift also increased for the aligned rhythms (in-phase) compared to rhythms with rhythmic subdivisions (anti-phase). Finally, the tested algorithmic models replicated the confederate's tempo drift without the use of any period correction mechanisms. This research advances our understanding of unintentional tempo drift, offering insights into ensemble dynamics and models of temporal coordination in groups. Implications for musical coordination and avenues for future research are discussed.

While Vector Symbolic Architectures (VSAs) are promising for modelling spatial cognition, their application is currently limited to artificially generated images and simple spatial queries. We propose VSA4VQA – a novel 4D implementation of VSAs that implements a mental representation of natural images for the challenging task of Visual Question Answering (VQA). VSA4VQA is the first model to scale a VSA to complex spatial queries. Our method is based on the Semantic Pointer Architecture (SPA) to encode objects in a hyperdimensional vector space. To encode natural images, we extend the SPA to include dimensions for object's width and height in addition to their spatial location. To perform spatial queries we further introduce learned spatial query masks and integrate a pre-trained vision-language model for answering attribute-related questions. We evaluate our method on the GQA benchmark dataset and show that it can effectively encode natural images, achieving competitive performance to state-of-the-art deep learning methods for zero-shot VQA.

Moral inference is an emerging topic of critical importance in artificial intelligence. The contemporary approach often relies on language modelling to infer moral relevance or moral properties of a concept such as "smoking". This approach demands complex parameterisation and costly computation, and it tends to disconnect with psychological accounts of moralization. We present a simple cognitive model for moral inference grounded in theories of moralization. Our model builds on word association network known to capture human semantics and draws on rich psychological data. We demonstrate that our moral association graph model performs competitively to state-of-the-art language models, where we evaluate them against a comprehensive set of data for automated inference of moral norms and moral judgment of concepts, and in-context moral inference. Moreover, we show that our model discovers intuitive concepts underlying moral judgment and is applicable to informing short-term temporal changes in moral perception.

Infants' bodies, brains, and environments are ever-changing. Although this continuous transformation is a fundamental feature of development, how infants actively adapt and learn amidst such volatility is still unknown. To address this, we devised a novel learning task in which the location of a reward was systematically altered, transitioning from stable to volatile periods. Through computational modelling, we inferred from the infants' gaze and pupil data the learning processes that enabled them to navigate these changing environments. We found that infants' tonic pupil size reflected trial-by-trial changes in the level of environmental volatility. Moreover, phasic changes in pupil size when observing the reward indicated that infants relied on the information about volatility to optimize their learning. This resulted in the successful performance of the task, as indicated by the pattern of anticipatory looks to the correct reward locations. Together, these results identify the active role that infants play in adapting to change.

It has long been assumed that infants' ability to discriminate between languages stems from their sensitivity to speech rhythm, i.e., organized temporal structure of vowels and consonants in a language. However, the relationship between speech rhythm and language discrimination has not been directly demonstrated. Here, we use computational modeling and train models of speech perception with and without access to information about rhythm. We test these models on language discrimination, and find that access to rhythm does not affect the success of the model in replicating infant language discrimination results. Our findings challenge the relationship between rhythm and language discrimination, and have implications for theories of language acquisition.

Choice deadlines are commonly imposed in decision-making research to incentivize speedy responses and sustained attention to the task settings. However, computational models of choice and response times routinely overlook this deadline, instead simply omitting trials past the deadline from further analysis. This choice is made under the implicit assumption that parameter estimation is not significantly affected by ignoring these omissions. Using new tools from likelihood-free inference, here we elucidate the degree to which omitting omissions, even in seemingly benign settings, can lead researchers astray. We explore the phenomenon using a Sequential Sampling Model (SSM) with collapsing boundaries as a test-bed.

Decisions about extinction risks are ubiquitous in everyday life and for our continued existence as a species. We introduce a new risky-choice task that can be used to study this topic: The Extinction Gambling Task. Here, we investigate two versions of this task: a Keep variant, where participants cannot accumulate any more earnings after the extinction event, and a Lose variant, where extinction also wipes out all previous earnings. We derive optimal solutions for both variants and compare them to behavioural data. Our findings suggest that people understand the difference between the two variants and their behaviour is qualitatively in line with the optimal solution. Further, we find evidence for risk-aversion in the Keep condition but not in the Lose condition. We hope that this task can facilitate further research on this vital topic.

Humans preferentially recall items that are presented in close temporal proximity together -- a phenomenon known as the "temporal contiguity effect". In this study, we investigate whether this phenomenon emerges naturally when training a recurrent neural network with episodic memory on free recall tasks. The model learns to recall items in the order they were presented, consistent with the human contiguity effect. The strength of this effect predicts the performance of individual networks, mirroring experimental findings in humans where stronger contiguity effects predict higher recall performance. The contiguity effect in the model is supported by a neural representation of item index, resembling the `method of loci'. This differs from prominent computational models of human memory, which use a slow decay of past information to guide sequential retrieval. Our findings provide insights into the mechanisms underlying episodic memory and pave the way for future studies of its interactions with other cognitive processes.

Current models of cognitive control frame its allocation as a process of expected utility maximization. The benefits of a candidate action are weighed against the costs of that control allocation (e.g. opportunity costs). Recent theorizing has found that it is normative to account for the value of learning when determining control allocation. Here, we sought to test whether learning expectations could explain people's initial control allocation in a standard dot-motion perceptual task. We found that subjects' initial skill level and learning rate in a first block were able to predict their initial willingness to accumulate evidence in a second block, interpreted as a greater control allocation for the task. Our findings support the hypothesis that agents consider the learnability of a task when deciding how much cognitive control to allocate to that task.

In term series problems where multiple mental models can be constructed, partial-order models can be created as mental representations, which make it easier to perceive the symmetry of the terms. To test these hypotheses, we categorized multi-model (indeterminate) term series problems according to the patterns of partial-order models that could be constructed, and analyzed the reasoning performance for each pattern. These results suggest that reasoners tend to use the symmetry of terms to reduce the cognitive load of reasoning. Analysis of the patterns of incorrect answers also suggests that attempts to exploit the symmetry of the term may be biased, leading to errors in reasoning.

Cognitive control is thought to regulate the conflict between stability---maintaining the current task in the face of distraction---and flexibility---switching to a new task of greater priority. However, evidence conflicts regarding when and to what extent stability and flexibility trade-off. A normative theory of flexibility and stability may help clarify when and why we should expect such trade-offs to occur. Towards such a theory, we model task-switching as a problem of decision-making under uncertainty, in which the decision-maker must simultaneously infer both the identity of a stimulus and the task governing the correct response to that stimulus. We find that optimal behavior is either extremely stable or extremely flexible, but not both, indicating a normative basis for a trade-off between the two. However, we also show that a sub-optimal but more realistic decision-maker exhibits behavior between these two extremes, and more closely resembles experimental data.

We conduct tests of a hybrid-similarity exemplar model on its ability to account for the context-dependent memorability of items embedded in high-dimensional category spaces. According to the model, recognition judgments are based on the summed similarity of test items to studied exemplars. The model allows for the idea that “self-similarity” among objects differs due to matching on highly salient distinctive features. Participants viewed a study list of rock images belonging to geologically defined categories where the number of studied items from each category was manipulated, and their old-new recognition performance was then tested. With a minimum of parameter estimation, the model provided good accounts of changing levels of memorability due to contextual effects of category size, within- and between-category similarity, and the presence of distinctive features. We discuss future directions for improving upon the current predictions from the model.