| Publications of year 2025 |
| Book chapters |
| Articles in journals |
| Abstract: | Premature birth impacts the development of superior temporal brain regions, including the superior temporal sulcus (STS), a key cortical area for language, voice recognition, and music processing. Using three distinct newborn imaging datasets, we examined the impact of premature birth on STS morphology at term-equivalent age. In the large cohort of the Developing Human Connectome Project (dHCP), we observed a linear relationship between gestational age at birth and STS depth, with earlier birth associated with a shallower STS. We hypothesized that this effect may have resulted from reduced structured auditory stimulation during a critical period of perisylvian network development. To test this hypothesis, we analyzed two additional published cohorts in which preterm neonates were exposed to contrasting auditory environments: either enhanced with structured music or minimized in quiet private rooms. We found that music exposure was associated with deeper STS, while a quieter environment was linked to further STS shallowing. Although the cross-sectional design limits causal inference, our findings suggest that early auditory experience--both in and ex utero--may influence the structural development of temporal brain regions. These results highlight the need to deepen our understanding of environmental influences in order to optimize postnatal settings that support the harmonious development of auditory and language networks. |
| Abstract: | Background and hypothesis Effort allocation is a crucial component of amotivation in schizophrenia. This study investigates the hypothesis that schizophrenia is associated with impairments in dynamic cost/benefit decision-making processes. Study design We employed a modified version of the effort allocation task developed by Meyniel et al. (2013). Participants were asked to allocate effort during 30-s intervals to maximize their gains. We examined the effects of task difficulty and incentive levels on participants' effort allocation on a trial-by-trial basis. Study results Individuals with schizophrenia (N = 25) showed decreased capacity to adapt dynamically to task parameters, as compared to healthy controls (N = 25). (1) Both populations increased the duration of each effort based on difficulty. Only healthy controls decreased rest duration based on incentive. The magnitude of these adaptations was significantly decreased in people with schizophrenia (difficulty: d = 1.25, incentive: d = 0.91). (2) Both groups decreased effort re-initiations with increasing difficulty with significant differences in the magnitude of adaptation between groups. (3) Participants with schizophrenia spent less time exerting effort above the required threshold, resulting in lower overall gains compared to healthy controls (?2 = 0.17). Conclusions Individuals with schizophrenia exhibit a selective impairment in effort-cost decision-making. This deficit may contribute to maladaptive behavior patterns characterized by suboptimal effort allocation and reduced goal-direct activities. |
| Abstract: | Number sense, the ability to rapidly estimate object quantities in a visual scene without precise counting, is a crucial cognitive capacity found in humans and many other animals. Recent studies have identified artificial neurons tuned to numbers of items in biologically inspired vision models, even before training, and proposed these artificial neural networks as candidate models for the emergence of number sense in the brain. But real-world numerosity perception requires abstraction from the properties of individual objects and their contexts, unlike the simplified dot patterns used in previous studies. Using novel synthetically generated photorealistic stimuli, we show that deep convolutional neural networks optimized for object recognition encode information on approximate numerosity across diverse objects and scene types, which could be linearly read out from distributed activity patterns of later convolutional layers of different network architectures tested. In contrast, untrained networks with random weights failed to represent numerosity with abstractness to other visual properties and instead captured mainly low-level visual features. Our findings emphasize the importance of using complex, naturalistic stimuli to investigate mechanisms of number sense in both biological and artificial systems, and they suggest that the capacity of untrained networks to account for early-life numerical abilities should be reassessed. They further point to a possible, so far underappreciated, contribution of the brain's ventral visual pathway to representing numerosity with abstractness to other high-level visual properties. |
| Abstract: | While deep learning has enabled the decoding of language from intracranial brain recordings, achieving this with non-invasive recordings remains an open challenge. We introduce a deep learning pipeline to decode individual words from electro- (EEG) and magneto-encephalography (MEG) signals. We evaluate our approach on seven public datasets and two datasets which we collect ourselves, amounting to a total of 723 participants reading or listening to five million words in three languages. Our model outperforms existing methods consistently across participants, devices, languages, and tasks, and can decode words absent from the training set. Our analyses highlight the importance of the recording device and experimental protocol: MEG and reading are easier to decode than EEG and listening, and decoding performance consistently increases with the amount of data used for training and for averaging during testing. Overall, our findings delineate the path and remaining challenges towards building non-invasive brain decoders for natural language. |
| Abstract: | Mathematics is an underexplored domain of human cognition. While many studies have focused on subsets of math concepts such as numbers, fractions, or geometric shapes, few have ventured beyond these elementary domains. Here, we attempted to map out the full space of math concepts and to answer two specific questions: can distributed semantic models, such a GloVe, provide a satisfactory fit to human semantic judgements in mathematics? And how does this fit vary with education? We first analyzed all of the French and English Wikipedia pages with math contents, and used a semi-automatic procedure to extract the 1000 most frequent math terms in both languages. In a second step, we collected extensive behavioral judgements of familiarity and semantic similarity between them. About half of the variance in human similarity judgements was explained by vector embeddings that attempt to capture latent semantic structures based on cooccurence statistics. Participants' self-reported level of education modulated familiarity and similarity, allowing us to create a partial hierarchy among high-level math concepts. Our results converge onto the proposal of a map of math space, organized as a database of math terms with information about their frequency, familiarity, grade of acquisition, and entanglement with other concepts. |
| Abstract: | Interest in statistical learning in developmental studies stems from the observation that 8-month-olds were able to extract words from a monotone speech stream solely using the transition probabilities (TP) between syllables (). A simple mechanism was thus part of the human infant's toolbox for discovering regularities in language. Since this seminal study, observations on statistical learning capabilities have multiplied across domains and species, challenging the hypothesis of a dedicated mechanism for language acquisition. Here, we leverage the two dimensions conveyed by speech -speaker identity and phonemes- to examine (1) whether neonates can compute TPs on one dimension despite irrelevant variation on the other and (2) whether the linguistic dimension enjoys an advantage over the voice dimension. In two experiments, we exposed neonates to artificial speech streams constructed by concatenating syllables while recording EEG. The sequence had a statistical structure based either on the phonetic content, while the voices varied randomly (Experiment 1) or on voices with random phonetic content (Experiment 2). After familiarisation, neonates heard isolated duplets adhering, or not, to the structure they were familiarised with. In both experiments, we observed neural entrainment at the frequency of the regularity and distinct Event-Related Potentials (ERP) to correct and incorrect duplets, highlighting the universality of statistical learning mechanisms and suggesting it operates on virtually any dimension the input is factorised. However, only linguistic duplets elicited a specific ERP component, potentially an N400 precursor, suggesting a lexical stage triggered by phonetic regularities already at birth. These results show that, from birth, multiple input regularities can be processed in parallel and feed different higher-order networks. |
| Abstract: | Assessing probabilities and predicting future events are fundamental for perception and adaptive behavior, yet the neural representations of probability remain elusive. While previous studies have shown that neural activity in several brain regions correlates with probability-related factors such as surprise and uncertainty, similar correlations have not been found for probability. Here, using 7 Tesla functional magnetic resonance imaging, we uncover a representation of the probability of the next event in a sequence within the human dorsolateral prefrontal and intraparietal cortices. Crucially, univariate and multivariate analyses revealed that this representation employs a highly non-monotonic code. Tuning curves for probability exhibit selectivity to various probability ranges, while the code for confidence accompanying these estimates is predominantly monotonic. Given such diversity in tuning curves, future studies should move from assuming monotonic or simple canonical forms of tuning curves to considering richer representations, and clarify why different types of code exist. |
| Abstract: | ABSTRACT At the physical level, the experience of pitch has a single determinant: the repetition rate of a waveform in the acoustic signal. Yet, psychologists describe pitch as composed of two perceptual dimensions, height and chroma. Chroma accounts for octave equivalence, whereby sounds with fundamental frequencies at a 1:2 ratio are perceived as sharing the same pitch. A current controversy debates whether chroma is a basic perceptual property dependent on biological constraints or a higher-order cognitive construct shaped by culture. Here, we used high-density electroencephalography (EEG) and time-resolved multivariate pattern analyses to characterize pitch processing in humans at 3 months of age. We found that, when exposed to repetitive sequences of orchestral tones, infants encode two separate pitch-related dimensions automatically and with divergent dynamics. Namely, our classifiers isolated height-specific information from the neural signal rapidly after the onset of the auditory sequences. Beyond approximately 600 ms, the performance of pitch height decoders fell to chance level and did not recover. In contrast, neural patterns displaying octave equivalence were retrieved later in the trial, over multiple time windows throughout the unfolding of the auditory sequence, and after sequence offset. Overall, this study reveals that very early in human development, the pitch of naturally rich tones is processed over two distinct encoding stages, capturing not only their absolute height but also their relative position in the octave. We speculate that separate encoding mechanisms reflect distinct functional roles carried by the two dimensions. |
| Abstract: | Visual numerosity, traditionally linked to the parietal cortex, is now thought to be represented across a broader cortical network, including early visual and associative areas in both streams. However, how numerosity is encoded relative to other visual features remains unclear. We conducted a whole-brain functional magnetic resonance imaging (fMRI) study with thirty-one adults performing a numerosity estimation task on visual sets varying in number, item size, total item area, field area, and density, ensuring tight stimulus control. Using model-based representational similarity analyses, we found numerosity represented independently of other visual properties in early visual areas and amplified in retinotopic and non-retinotopic associative regions across both streams. Dimensionality reduction of BOLD patterns revealed distinct geometries: a one-dimensional representation of numerical rank in early visual and ventral retinotopic areas, and a curved structure encoding rank and distance-to-endpoints in associative dorsal and ventral regions. These results demonstrate distinct neural coding schemes for numerosity across cortical regions. |
| Abstract: | Inspired by recent discoveries of neural populations that track time for specific moments (time cells) and elapsed durations (temporal context and periodic time cells), this review, based on a minisymposium presented at the 2025 annual meeting of the Society for Neuroscience, brings together macro-, meso-, and micro-scale neural evidence to formulate hypotheses about how episodic time--the tracking and organizing of events in time as we experience, store, and retrieve them from memory--is coded and processed in the mammalian brain. We also discuss computational principles and relationships to other related phenomena such as memory replay and emotional states. |
| Abstract: | The superior temporal sulcus (STS) plays a central role in auditory and linguistic processing and undergoes rapid development during the last trimester of gestation. Yet, the extent to which its development is shaped by early sensory experience remains unclear. Premature birth offers a unique opportunity to address this question, as it exposes the brain to an extra-uterine auditory environment at a critical stage of network maturation. We analyzed resting-state fMRI data in 116 neonates (63 males), scanned at term-equivalent age but born at varying gestational age (24.3 to 41.7 weeks gestational age) using the developing Human Connectome Project (dHCP) database. Functional connectivity was computed in native space using regions of interest based on each infant's sulcal anatomy to assess the respective contributions of STS subregions. Our analyses reveal a functional division between the inferior and superior banks of the STS, with the inferior bank showing stronger connectivity to distant parietal and frontal areas along the dorsal language pathway. The left posterior STS emerged as a functional hub, displaying broad inter-area connectivity. Longer gestations correlated with increased local connectivity, notably in the right temporal region, despite equal age at scan. Additionally, female neonates exhibited stronger connectivity from the left posterior STS compared to males. These findings highlight the early emergence of adult-like auditory-linguistic networks and their sensitivity to the in-utero environment. Further research is needed to investigate the consequences of these early differences and to determine which postnatal interventions might help compensate, if necessary. |
| Abstract: | Around the world, teenage boys outperform girls on mathematics tests, and men are more likely to pursue related careers -- despite baby boys showing no superior sense of numbers or grasp of logic. Now, a gigantic study of schoolchildren in France pinpoints that this 'mathematical gender gap' appears during the first year of school. The finding could help to focus efforts to stop girls from falling behind. Boys and girls receive similar maths scores at the start of school, but boys pull ahead of girls after just four months (see 'Watch the mathematics gender gap emerge'). A more dramatic gap in mathematical performance emerges after 12 months of school, according to the analysis, published on 11 June in Nature1. "This paper suggests that the gender inequalities in children's maths performance aren't innate or inevitable," says psychologist Jillian Lauer at the University of Cambridge, UK. "If we want to stop girls from falling behind, we need to focus on their early experiences at school." |
| Abstract: | Abstract Postdiction is a perceptual phenomenon where the perception of an earlier stimulus is influenced by a later one. This effect is commonly studied using the 'rabbit illusion', in which temporally regular, but spatially irregular, stimuli are perceived as equidistant. While previous research has focused on short inter-stimulus intervals (100-200 ms), the role of longer intervals, which may engage late attentional processes, remains unexplored. This study investigates whether postdiction is purely perceptual or also involves attentional mechanisms by using visual stimuli separated by extended intervals. 33 participants (17 females) were assigned to two experimental groups with two different temporal inter-flash intervals (IFI) between stimuli (250 ms: 250-IFI group; 500 ms: 500-IFI). Two stimulation protocols of active transcranial electrical stimulation (tES) and one control condition were tested on the left precuneus/inferior parietal gyrus: (i) transcranial alternating current stimulation (tACS) at the individual alpha frequency (IAF) (IAF-tACS); (ii) transcranial random noise stimulation across the whole alpha band (i.e., 8-12 Hz, Alpha-tRNS) and (iii) a placebo (Sham) stimulation. The postdiction phenomenon was observable in both experimental groups. The participants in the 500-IFI group demonstrated enhanced performance in detecting the illusion during the rabbit illusion task when IAF-tACS was applied. The behavioral results suggest that attentional functions, beyond perceptual ones, play a key role in the postdiction phenomenon. |
| Abstract: | Abstract Temporal landmarks are salient events that structure the way humans think about time. They may be personal events, such as one's birthday, or shared cultural events, such as the COVID-19 pandemic. Due to societal habits, the cyclical weekly structure - for example, working on weekdays, resting on the weekends - helps individuals orient themselves in time. In the "day-of-the-week effect," individuals are faster at reporting which day of the week it is on weekends than they are on weekdays. Herein, we hypothesized that the disruption of social habits during the COVID-19 pandemic lockdowns may have weakened this effect, thereby accounting for the "Blursday" phenomenon. In the current study, speeded responses to the question "What day of the week is it?" were collected online from 1,742 French participants, during and after the lockdown periods. We found that reaction times for days of the weekends remained faster than for weekdays during the lockdown, although the overall reaction times were significantly slower during lockdown. We also found that responses were slower as governmental stringency rules and restrictions in mobility increased. Our results suggest that the weekend landmark remains a stable temporal anchor in French culture despite the experienced temporal distortions induced by the disruption of social habits during the pandemic. We conclude that cultural temporal landmarks shape socially shared temporal cognitive maps. |
| Conference proceedings |
| Abstract: | Maintaining accurate beliefs in a changing and noisy environment is a challenging computational problem. Previous studies have shown that humans adapt their learning dynamically, especially in the face of change. This conclusion is mostly supported in the context of magnitude learning (e.g., tracking a reward amount, an object position), and currently remains more uncertain in the case of probability learning (e.g., tracking the probability of an event occurring). Here, we initiate an open benchmarking approach to uncover the computations humans use for probability learning. We compared a wide range of models--including optimal Bayesian models, suboptimal variants, and simple prediction error-based update rules, using several datasets in which participants provided trial-by-trial probability estimates. Bayesian inference often outperformed simple prediction error-based models, despite being more computationally demanding and often considered less biologically plausible. Furthermore, inference strategies appear to depend on environmental volatility: under moderate volatility, an optimal Bayesian model best explains behavior, whereas in more stable environments, a simpler Bayesian approximation is better. These results so far highlight the sophistication of human adaptive learning for probability and suggest that humans can adapt their inference strategies based on environmental context. We invite others to contribute models and datasets to this benchmark to refine these conclusions. |
This material is presented to ensure timely dissemination of scholarly and technical work. Copyright and all rights therein are retained by authors or by other copyright holders. All person copying this information are expected to adhere to the terms and constraints invoked by each author's copyright. In most cases, these works may not be reposted without the explicit permission of the copyright holder.
Note that this is not the exhaustive list of publications, but only a selection. Contact the individual authors for complete lists of references.
This document was translated from BibTEX by bibtex2html