Digital platforms generate unprecedented traces of human behaviour, offering new methodological approaches to understanding collective action, polarisation, and social dynamics. Through analysis of millions of digital traces across multiple studies, we demonstrate how online behaviours predict offline action: Brexit-related tribal discourse responds to real-world events, machine learning models achieve 80% accuracy in predicting real-world protest attendance from digital signals, and social validation through "likes" emerges as a key driver of mobilization. Extending this approach to conspiracy narratives reveals how digital traces illuminate psychological mechanisms of belief and community formation. Longitudinal analysis of YouTube conspiracy content demonstrates how narratives systematically address existential, epistemic, and social needs, while examination of alt-tech platforms shows how emotions of anger, contempt, and disgust correlate with violence-legitimating discourse, with significant differences between narratives associated with offline violence versus peaceful communities. This work establishes digital traces as both methodological innovation and theoretical lens, demonstrating that computational social science can illuminate fundamental questions about polarisation, mobilisation, and collective behaviour across contexts from electoral politics to conspiracy communities.

Recent research has seen many behavioral comparisons between humans and deep neural networks (DNNs) in the domain of image classification. Often, comparison studies focus on the end-result of the learning process by measuring and comparing the similarities in the representations of object categories once they have been formed. However, the process of how these representations emerge—that is, the behavioral changes and intermediate stages observed during the acquisition—is less often directly and empirically compared. In this talk, I'm going to report a detailed investigation of the learning dynamics in human observers and various classic and state-of-the-art DNNs. We develop a constrained supervised learning environment to align learning-relevant conditions such as starting point, input modality, available input data and the feedback provided. Across the whole learning process we evaluate and compare how well learned representations can be generalized to previously unseen test data. Comparisons across the entire learning process indicate that DNNs demonstrate a level of data efficiency comparable to human learners, challenging some prevailing assumptions in the field. However, our results also reveal representational differences: while DNNs' learning is characterized by a pronounced generalisation lag, humans appear to immediately acquire generalizable representations without a preliminary phase of learning training set-specific information that is only later transferred to novel data.

People largely differ with respect to how well they can learn, memorize, and perceive faces. In this talk, I address two potential sources of variation. One factor might be people’s ability to adapt their perception to the kind of faces they are currently exposed to. For instance, some studies report that those who show larger adaptation effects are also better at performing face learning and memory tasks. Another factor might be people’s sensitivity to perceive fine differences between similar-looking faces. In fact, one study shows that the brain of good performers in a face memory task shows larger neural differences between similar-looking faces. Capitalizing on this body of evidence, I present a behavioural study where I explore the relationship between people’s perceptual adaptability and sensitivity and their individual face processing performance.

Frequency tagging based on fast periodic visual stimulation (FPVS) provides a window into ongoing visual and cognitive processing and can be leveraged to measure rule learning and high-level categorization. In this talk, I will present data demonstrating highly proficient categorization as living and non-living in preschool children, and characterize the development of this ability during infancy. In addition to associating cognitive functions with development, an intriguing question is whether frequency tagging also captures enduring individual differences, e.g. in general cognitive abilities. First studies indicate high psychometric quality of FPVS categorization responses (XU et al., Dzhelyova), providing a basis for research on individual differences. I will present results from a pilot study demonstrating high correlations between FPVS categorization responses and behavioral measures of processing speed and fluid intelligences. Drawing upon this first evidence, I will discuss the potential of frequency tagging for diagnosing cognitive functions across development.

This talk aims to highlight the immense potential of Artificial Intelligence (AI) in advancing the field of psychology and cognitive neuroscience. Through the integration of machine learning algorithms, big data analytics, and neuroimaging techniques, AI has the potential to revolutionize the way we study human cognition and brain characteristics. In this talk, I will highlight our latest scientific advancements in utilizing AI to gain deeper insights into variations in cognitive performance across the lifespan and along the continuum from healthy to pathological functioning. The presentation will showcase cutting-edge examples of AI-driven applications, such as deep learning for automated scoring of neuropsychological tests, natural language processing to characeterize semantic coherence of patients with psychosis, and other application to diagnose and treat psychiatric and neurological disorders. Furthermore, the talk will address the challenges and ethical considerations associated with using AI in psychological research, such as data privacy, bias, and interpretability. Finally, the talk will discuss future directions and opportunities for further advancements in this dynamic field.

Visual working memory (VWM) is limited in capacity, but memorizing meaningful objects may refine this limitation. However, meaningless and meaningful stimuli usually differ perceptually and an object’s association with meaning is typically already established before the actual experiment. We applied a strict control over these potential confounds by asking observers (N=45) to actively learn associations of (initially) meaningless objects. To this end, a change detection task presented Chinese characters, which were meaningless to our observers. Subsequently, half of the characters were consistently paired with pictures of animals. Then, the initial change detection task was repeated. The results revealed enhanced VWM performance after learning, in particular for meaning-associated characters (though not quite reaching the accuracy level attained by N=20 native Chinese observers). These results thus provide direct experimental evidence that the short-term retention of objects benefits from active learning of an object’s association with meaning in long-term memory.

The ‘quantification’ of perception is arguably both one of the most important and most difficult aspects of perception study. This is particularly true in visual perception, in which the evaluation of the perceptual threshold is a pillar of the experimental process. The choice of the correct adaptive psychometric procedure, as well as the selection of the proper parameters, is a difficult but key aspect of the experimental protocol. For instance, Bayesian methods such as QUEST, require the a priori choice of a family of functions (e.g. Gaussian), which is rarely known before the experiment, as well as the specification of multiple parameters. Importantly, the choice of an ill-fitted function or parameters will induce costly mistakes and errors in the experimental process. In this talk we discuss the existing methods and introduce a new adaptive procedure to solve this problem, named, ZOOM (Zooming Optimistic Optimization of Models), based on recent advances in optimization and statistical learning. Compared to existing approaches, ZOOM is completely parameter free and model-free, i.e. can be applied on any arbitrary psychometric problem. Moreover, ZOOM parameters are self-tuned, thus do not need to be manually chosen using heuristics (eg. step size in the Staircase method), preventing further errors. Finally, ZOOM is based on state-of-the-art optimization theory, providing strong mathematical guarantees that are missing from many of its alternatives, while being the most accurate and robust in real life conditions. In our experiments and simulations, ZOOM was found to be significantly better than its alternative, in particular for difficult psychometric functions or when the parameters when not properly chosen. ZOOM is open source, and its implementation is freely available on the web. Given these advantages and its ease of use, we argue that ZOOM can improve the process of many psychophysics experiments.

Coregraph is a tool under development that allows us to collect high-density data patterns during the administration of classic neuropsychological tests such as the Trail Making Test and Clock Drawing Test. These tests are widely used to evaluate cognitive function and screen for neurodegenerative disorders, but traditional methods of data collection only yield sparse information, such as test completion time or error types. By contrast, the high-density data collected with Coregraph may contribute to a better understanding of the cognitive processes involved in executing these tests. In addition, Coregraph may potentially revolutionize the field of cognitive evaluation by aiding in the prediction of cognitive deficits and in the identification of early signs of neurodegenerative disorders such as Alzheimer's dementia. By analyzing high-density graphomotor data through techniques like manual feature engineering and machine learning, we can uncover patterns and relationships that would be otherwise hidden with traditional methods of data analysis. We are currently in the process of determining the most effective methods of feature extraction and feature analysis to develop Coregraph to its full potential.

I describe a well-tested Python module that does automated alignment and warping of faces images, and some advantages over existing solutions. An additional tool I’ve developed does automated extraction of facial features, which can be used in a number of interesting ways. I illustrate the value of wavelet-based features with a brief description of 2 recent studies: perceptual in-painting, and the robustness of the whole-part advantage across a large stimulus set. Finally, I discuss the suitability of various deep learning models for generating stimuli to study perceptual face spaces. I believe those interested in the forensic aspects of face perception may find this talk useful.

In this lecture, I would like to share with the broad audience the empirical results gathered and the theoretical advancements made in the framework of the Lendület project entitled ’The cognitive basis of human sociality’. The main objective of this project was to understand the mechanisms that enable the unique sociality of humans, from the angle of cognitive science. In my talk,  I will focus on recent empirical evidence in the study of three fundamental social cognitive functions (social categorization, theory of mind and social learning; mainly from the empirical lenses of developmental psychology) in order to outline a theory that emphasizes the need to consider their interconnectedness. The proposal is that the ability to represent the social world along categories and the capacity to read others’ minds are used in an integrated way to efficiently assess the epistemic states of fellow humans by creating a shared representational space. The emergence of this shared representational space is both the result of and a prerequisite to efficient learning about the physical and social environment.

The literature suggests that ensemble coding, the ability to represent the gist of sets, may be an underlying mechanism for becoming familiar with newly encountered faces. This phenomenon was investigated by introducing a new training paradigm that involves incidental learning of target identities interspersed among distractors. The effectiveness of this training paradigm was explored in Study 1, which revealed that unfamiliar observers who learned the faces incidentally performed just as well as the observers who were instructed to learn the faces, and the intervening distractors did not disrupt familiarization. Using the same training paradigm, ensemble coding was investigated as an underlying mechanism for face familiarization in Study 2 by measuring familiarity with the targets at different time points using average images created either by seen or unseen encounters of the target. The results revealed that observers whose familiarity was tested using seen averages outperformed the observers who were tested using unseen averages, however, this discrepancy diminished over time. In other words, successful recognition of the target faces became less reliant on the previously encountered exemplars over time, suggesting an exemplar-independent representation that is likely achieved through ensemble coding. Taken together, the results from the current experiment provide direct evidence for ensemble coding as a viable underlying mechanism for face familiarization, that faces that are interspersed among distractors can be learned incidentally.

The ability to recognize someone’s voice spans a broad spectrum with phonagnosia on the low end and super recognition at the high end. Yet there is no standardized test to measure the individual ability to learn and recognize newly-learnt voices with samples of speech-like phonetic variability. We have developed the Jena Voice Learning and Memory Test (JVLMT), a 20 min-test based on item response theory and applicable across different languages. The JVLMT consists of three phases in which participants are familiarized with eight speakers in two stages and then perform a three-alternative forced choice recognition task, using pseudo sentences devoid of semantic content. Acoustic (dis)similarity analyses were used to create items with different levels of difficulty. Test scores are based on 22 Rasch-conform items. Items were selected and validated in online studies based on 232 and 454 participants, respectively. Mean accuracy is 0.51 with an SD of .18. The JVLMT showed high and moderate correlations with convergent validation tests (Bangor Voice Matching Test; Glasgow Voice Memory Test) and a weak correlation with a discriminant validation test (Digit Span). Empirical (marginal) reliability is 0.66. Four participants with super recognition (at least 2 SDs above the mean) and 7 participants with phonagnosia (at least 2 SDs below the mean) were identified. The JVLMT is a promising screen too for voice recognition abilities in a scientific and neuropsychological context.