npj Science of Learning

In the context of education, attention can be considered the gateway for learning, yet it remains unclear which neural mechanisms of attention identified under controlled laboratory conditions are most relevant when children engage in meaningful learning. Here, we addressed this question by experimentally manipulating attention while 5th- and 6th-grade students learned novel educational content from their own teacher. Working in partnership with an experienced classroom teacher, we co-developed naturalistic auditory and visual learning streams and manipulated whether students prioritized or ignored the speech. Using school-based electroencephalography and temporal response function modeling, we examined whether attention modulated early sensory or later stages of cortical speech processing. Attention selectively modulated speech processing at approximately 170 ms, with no evidence for modulation at earlier sensory stages, supporting a predominant role for late-stage attentional selection during learning. Importantly, individual differences in attentional modulation were associated with learning: students who more strongly increased neural tracking of the speech when it was task-relevant learned more effectively from spoken instruction. The same late-stage neural mechanism also distinguished students whom their teacher independently identified as stronger attenders during everyday classroom learning. Together, these findings connect late-stage attentional modulation across experimental neural dynamics, individual learning outcomes, and teacher observations of classroom behavior. More broadly, they demonstrate how studying attention within educationally meaningful contexts can help identify which neural mechanisms are most consequential for successful learning.

Learning depends not only on the content of what we learn, but also on how we learn and on how experiences are structured over time. To investigate how task similarity and training regime interact during learning, we trained participants on spatial and conceptual learning tasks that shared either similar or distinct underlying structures, using either interleaved or blocked regimes. Interleaving the two tasks hindered performance when their structures were similar, compared to when they were different. In contrast, blocked training produced the opposite effect: it improved performance and facilitated transfer across similar tasks. This effect, however, emerged only when participants first learned the conceptual task, followed by the spatial task, suggesting an asymmetric interaction between task order and structural similarity. We also replicated our results using a neural network model, providing converging evidence for the computational principles governing the interplay between training regime and structural similarity in multi-task learning.

This study examined whether emotional valence modulates face working memory precision in adolescents with specific learning disorders (SLD) and whether mnemonic distortions differ across subtypes. Fifty-four adolescents with SLD (dyslexia: n = 19; dysgraphia: n = 18; dyscalculia: n = 17) and 42 typically developing peers completed a morphed delayed reproduction task using sad, neutral, and happy faces. Using a von Mises + uniform mixture model, we decomposed recall errors into precision, systematic bias, and guessing. The SLD group made significantly larger errors (r = 8722;0.61) and responded more slowly (r = -0.33). Mixture modeling showed that deficits arose from reduced precision ({kappa}: TD M = 8.72 vs. SLD M = 4.15; r = 0.54) and stronger systematic bias toward neutrality (: TD M = -0.66 vs. SLD M = -4.15; r = 0.34), with no differences in random guessing. Valence flattening was consistently larger in SLD. Among subtypes, dyscalculia showed the greatest working memory impairment, yet all three exhibited the same valence flattening pattern. Thus, emotional faces disrupt face working memory in SLD through reduced precision and systematic affective distortion. The bias parameter from mixture modeling may offer a clinically useful metric for subtype-sensitive assessment.

IntroductionExecutive functions (EF) are higher-order cognitive processes essential for academic performance in university settings. Although there is extensive research on EF training in children, studies in young adults are scarce, particularly those involving interventions tailored to specific needs. ObjectiveTo evaluate the effect of the University Executive Network-Training Program (NExT-U), based on specific needs, on the executive functioning of Cuban university students. MethodologyA quasi-experimental study with a non-equivalent control group and pretest-posttest measurements. Participants were 27 second-year Psychology students (74% female; mean age = 19 years). The experimental group (n=7) received three training sessions focused on Conscious Regulation of Behavior, Decision-Making, Emotional Regulation, and Monitoring of Responsibilities, identified through an initial assessment using the UEF-1 Scale. The control group (n=20) continued with their usual academic activities. Non-parametric analyses and the residual gain method were employed. ResultsThe experimental group showed significant improvements in Conscious Regulation of Behavior (p = .026; r = .51), Emotional Regulation (p = .030; r = .49), and the Supervisory Attention System (p = .046; r = .44), with large effect sizes. The control group experienced no significant changes in any of the functions evaluated. ConclusionsA brief, personalized program can enhance specific executive functions in university students, demonstrating cognitive plasticity in young adults. The findings support the design of contextually relevant interventions to strengthen transversal competencies in higher education.

Computer mouse weight has been proposed to influence human motor performance, but existing evidence comes largely from small, short-duration controlled studies, leaving its relationship to performance during naturalistic gameplay unresolved. We analyzed one-week skill improvement in Aimlabs, a popular first-person shooter aim-training and assessment platform, among 9,067 players whose telemetry recorded a mouse-model change (switchers) during self-directed game use, using a 21-day baseline and 7-day follow-up; 4,991 players with a single mouse model (non-switchers) observed over matched pseudo-anchor intervals provided a secondary contextual comparison. Post-switch mouse weight was represented by eight bands spanning 35-124 g. The primary multivariable generalized additive regression model accounted for baseline skill, pre- and post-switch practice intensity, and pre- and post-switch mouse sensitivity, and allowed the association between weight band and improvement to vary with baseline skill, post-switch practice, and post-switch sensitivity. For each weight band, predicted mean one-week improvement was standardized to the observed distribution of these covariates among switchers and reported in percentile-rank points. Statistical significance was evaluated using 1,000 permutations in which observed mouse weights were randomly reassigned among switchers without replacement and the model was refitted to generate empirical null distributions for the weight-related test statistics. Predicted improvement differed significantly across the eight weight bands (p = 0.001). Pairwise comparisons identified significant differences between bands (p < 0.05), but no single band was associated with significantly greater predicted improvement than all seven others. In this large naturalistic cohort, predicted improvement was highest for players with 35-44 g mice, followed by 45-54 and 55-64 g, and several pairwise comparisons showed significantly greater improvement in lighter than in heavier weight bands. Across the examined baseline-skill profiles, the lowest predicted improvements occurred in either the 85-99 or 115-124 g band. Predicted improvement differed across weight bands at the 25th-percentile and median baseline-skill values (p = 0.002 and p = 0.003) but not at the 75th percentile (p = 0.153). However, the weight-band x baseline-skill interaction was not significant (p = 0.181), indicating that the analysis did not establish that weight-related differences varied by skill level.

You remember what you attend to and you attend to what you remember. Memory is supported through engagement of mnemonic brain states, multivariate whole brain activity patterns that modulate down- stream processing and behavior. Attention can be focused externally to the environment or internally to thoughts and mental representations. To understand and promote successful cognition, it is critical to establish the extent to which the same brain states support both memory and attention processes. We recorded scalp EEG during three tasks in which we manipulated external and internal attention demands. We applied an independently-validated mnemonic state classifier to these data and find evidence in sup- port of our hypothesis that memory encoding and retrieval states map onto the external/internal axis of attention. Furthermore, our findings reveal sub-second fluctuations in mnemonic states. These results demonstrate that domain general mnemonic states support attentional orienting and can be used to detect moment-to-moment shifts in attention.

Self-directed learning allows learners to actively control their learning experience and has been shown to enhance memory compared with matched yoked learning. However, it remains unclear when and how active control modulates memory-related neural activity during learning and retrieval. We recorded electroencephalography (EEG) while participants encoded objects under active and yoked learning conditions and again during a delayed recognition test approximately 24 h later. We examined event-related potential (ERP) activity across earlier processing windows, including pre-stimulus slow potentials and early N2 activity, and later processing windows, including P300, late slow-wave, and post-stimulus slow-potential activity. We also examined encoding-retrieval similarity (ERS) between neural patterns during encoding and retrieval. Behaviorally, active control improved delayed recognition, with the advantage selectively expressed in detailed recognition. In the ERP analyses, earlier processing windows showed memory-related effects, with pre-stimulus slow-potential and N2 activity differentiating subsequently remembered from forgotten items, but were not modulated by active control. By contrast, active control modulated later memory-related ERP activity, with remembered-forgotten differences expressed during late stimulus-related and immediate post-stimulus processing only in the active condition. ERS showed a similar active-control modulation: memory-related encoding-retrieval pattern similarity was evident under active learning, but not under yoked learning, with this effect involving relatively late encoding and retrieval windows. Together, these findings support a constructive-processing account, suggesting that memory formation under active control depends more strongly on rich, detailed encoding representations that can be reinstated during retrieval.

Motor sequence performance improves during and between brief practice bouts (micro-online and offline gains). We compared both metrics across two groups: one exposed to an implicit motor sequence, and one not. Micro-online gains drove sequence-specific learning and positively correlated with overall skill. However, micro-offline gains were comparable between groups and did not track sequence-specific learning. We conclude that implicit motor sequence learning is driven by micro-online rather than micro-offline gains.

Introduction: The developing neural circuitry supporting adolescent risk-taking is vulnerable to metals, yet few studies have examined how early-life metal mixture exposure affects neural mechanisms underlying risk-taking. We used novel dentine biomarkers to identify critical windows of early-life metal mixture exposure associated with neurocognitive correlates of risky decision-making in early adolescence. Methods: Participants (ages 8-14, 45% female) were recruited from the Programming Research in Obesity, Growth, Environment and Social Stressors (PROGRESS) longitudinal birth cohort in Mexico City. Weekly concentrations of barium (Ba), copper (Cu), lithium (Li), magnesium (Mg), manganese (Mn), lead (Pb), tin (Sn), strontium (Sr), and zinc (Zn) were measured in children's deciduous teeth from 19 weeks pre-birth to 43 weeks postnatally. Neural activity during risky decision-making and reward processing was assessed using a functional magnetic resonance imaging gambling paradigm. Risk-taking tendency, risk sensitivity, and reward sensitivity were estimated from participants' decisions. We used lagged weighted quantile sum regression to examine time-varying associations between metal mixture exposure and neurocognitive outcomes, adjusting for child age, child sex, and socioeconomic status. Results: Higher metal mixture exposure during the early to late postnatal period was associated with increased risk-taking tendency and risk sensitivity. Exposure during the prenatal and late postnatal periods was associated with increased reward sensitivity and reduced activation in the angular gyrus/inferior parietal lobule during risky decision-making. Exposure during the late postnatal period was associated with increased activation in the insula/rolandic operculum during reward processing. Postnatal associations for risk-related and reward-related neural activation were primarily driven by Sr and Mg respectively, while postnatal associations for risk-taking tendency, risk sensitivity, and reward sensitivity were driven by multiple metals in the mixture. Conclusions: Early-life metal mixture exposure is differentially associated with neurocognitive correlates of risky decision-making in adolescence, depending on exposure timing.

Attention is foundational to learning, yet the extent to which features of the instructional environment differentially shape attentional engagement is not well understood. Here we leveraged mobile EEG and video-coded behavioral observation to examine attentional engagement in 6 to 10 years old children with and without a diagnosis of ADHD across instructional conditions varying in delivery modality (video watching, online, in-person) and management of learning (teacher-led versus student-led). EEG measures, including alpha-band (8-12Hz) oscillations, spectral slope and offset, and behavioral measures of active engagement (AE%), passive engagement (PE%), fidgeting, off-task behavior, were examined as a function of instructional context. Attentional engagement, as indicated by higher AE%, lower alpha power, flatter spectral slope and lower offset, was greatest in student-led learning, followed by teacher-led in person learning, synchronous online learning, and asynchronous learning, respectively. Child by instructional context interactions revealed that patterns of attentional engagement were largely consistent regardless of diagnosis, with group effects observed for motor behaviors (PE%, fidgeting) but not for measures of visual attention. Similarly, age showed only main effects, whereby older children showed higher passive engagement, lower fidgeting and off-task behavior, flatter spectral slope, and lower offset. The results underscore: (i) the importance of considering environmental context, such as instructional modality and management of learning, when examining mechanisms of attention and individual differences therein, and (ii) the value of recording multiple behavioral and neural measures, as not all indices commonly assumed to reflect attention capture the same underlying processes.

Motivational incentives are known to improve working memory (WM) when items differ in value, but whether they enhance performance when all items are equally valued, and if so, by what mechanism, remains unclear. Improvements could reflect enhanced memory-based representations or speed-accuracy trade-offs, whereby incentives promote slower, more cautious responding without improving memory itself. Here, we tested these alternatives across two experiments using an auditory WM task in which incentives varied across trials, but all items on a trial were equally incentivised. In Experiment 1,we allowed participants to resubmit their responses if desired. Reward-related improvements in WM performance were selectively observed when memory load was high, and only at serial positions associated with greater interference. These improvements were also dissociated from systematic response slowing and resubmission behaviour. This indicated that the benefits of incentives were unlikely to be attributable to response caution. Critically, in Experiment 2, motivational benefits persisted under strict response deadlines, where speed cannot be traded for accuracy. Computational modelling showed that reward selectively improved the precision of memory-based responses without reducing guessing. Together, these findings indicate that motivational incentives improve WM by enhancing the quality of memory-based representations, rather than by strategic slowing or altering response strategies.

Humans efficiently learn the temporal structure of speech, yet the underlying cognitive mechanisms remain unclear. Recent research in visuospatial sequential memory has proposed the successor representation (SR), important in reinforcement learning, which encodes multi-step transitional relationships. To test whether SR also supports the structural learning of rapidly changing speech syllable sequences, we adapted a pyramid-structured sequential learning paradigm using syllable streams presented at a natural speech rate (~ 4 Hz). Participants exhibited robust structural learning, evidenced by asymmetric memory performance modulated by adjacent transitional probabilities. Notably, the SR model captured recognition performance for non-adjacent syllable pairs. Moreover, SR-based learning effects generalized across all syllable sequences, regardless of adjacency. Furthermore, the effectiveness of SR learning was associated with cognitive flexibility and working memory capacity. Together, these findings suggest that humans rely on effortful SR-like mechanisms for the structural learning of syllable sequences, highlighting a modality-general computational principle for sequential learning across various temporal modulation scales.

Collective social events often require coordinated action by one group to become shared experience in another, yet how this transformation is organized across multiple brains remains unclear. Live music provides a tractable model of this problem because ensemble coordination, performer-audience alignment, and shared audience integration unfold within the same event. Here we tested whether performer-audience neural coupling acts as a cross-role neural interface linking ensemble coordination to shared audience integration. We recorded brain activity from a nine-person live performance system, consisting of a fixed three-performer ensemble and four independent audience groups, using synchronized multi-device functional near-infrared spectroscopy hyperscanning across live trio performance sessions. Inter-brain neural coupling was analyzed across three relational layers: performer-performer (PP), performer-audience (PA), and audience-audience (AA) coupling. Behavioral ratings showed strong affective engagement and shared evaluative alignment. Neural coupling during live performance was not expressed as a diffuse increase across channel pairs, but was organized into task-sensitive relational components with interpretable PC-corr network modules. Crucially, path-based mediation analyses revealed that PA components statistically bridged PP coordination and AA coupling, yielding multiple complete and partial PP [-&gt;] PA [-&gt;] AA pathways. Brain-behavior analyses further suggested that mediation-related PA components were linked to shared performance evaluation and emotional alignment. These findings identify performer-audience coupling as a cross-role neural interface through which coordinated production becomes linked to shared collective reception. Significant StatementHow coordinated actions become shared experiences is a major problem in social neuroscience, yet most studies examine pairs of people. Using live music as a model of group interaction, we found that brain-to-brain alignment was organized across social roles rather than arising as a uniform response to the same event. Performer-audience alignment occupied a bridging position between coordination within the ensemble and integration within the audience. This identifies a systems-level architecture through which collective action may become linked to collective experience. The framework moves multi-brain research beyond dyads and offers a general approach for studying classrooms, public speaking, theater, rituals, and team events, where one group generates structured behavior that another group jointly receives and interprets.

Reflecting how we organize visual experience in everyday memory, visual working memory is increasingly understood as a system in which individual item representations are organized within structures rather than maintained in isolation. Among these, relational structure may be especially consequential because, by specifying how one item value lies relative to another within a feature space, it could allow information about one remembered value to constrain which values are plausible for the other. Yet demonstrating such constraint is challenging because item-specific mnemonic evidence and relational evidence ordinarily support essentially the same estimate. We broke this equivalence with biased post-encoding feedback for one item, making item-specific and relation-based predictions for the other diverge. Across three experiments, participants remembered two sequentially presented orientations, with feedback for one shifted slightly clockwise or counterclockwise from its actual value. Participants incorporated this bias into memory for the feedback-provided orientation; critically, it also appeared in reports of the other orientation, which received no feedback, in the direction predicted by the signed angular offset linking the two remembered values. This feedback transfer weakened with increasing angular separation but occurred in both directions between the first and second orientations. These findings show that relational structure directly constrains individual value estimates in visual working memory, even for items encountered separately. By dissociating normally coincident item-specific and relation-based predictions, our approach reveals an otherwise hidden relational contribution. A probabilistic account explains these findings through joint inference from uncertain item-specific and relational evidence, with their relative uncertainties governing transfer strength.

Understanding the cognitive architecture of the human language faculty requires exploring the boundaries of both predisposition and environmental experience. However, previous research on extraordinary multilingualism has often confounded language aptitude with multilingual experience, obscuring their distinct neural correlates. Here, we leveraged a linguistically diverse sample (N=121) and extensive behavioural testing to dissociate language aptitude from multilingual experience, modelling both dimensions continuously in whole-brain speech processing. Language aptitude and multilingual experience were weakly related, and their dissociation was also evident at the neural level. Higher language aptitude showed a neural signature of efficiency, characterised by lower activation in core perisylvian regions. In contrast, higher multilingualism was associated with greater engagement of regions implicated in narrative, multimodal, and memory processing, and with recruitment of traditional language hubs only during degraded speech processing, likely reflecting active attempts to decode unintelligible input. Finally, aptitude and experience interacted within sensorimotor regions. Continuous quantification of multilingual experience proved more sensitive than artificial grouping. By disentangling language aptitude from multilingual experience, this work provides a more precise account of the multilingual brain, and shows that its neurobiology can be better understood by modelling predisposition and experience as distinct but interacting dimensions.

Task demand has a long-standing history in experimental psychology, with numerous investigations of the capacity of different systems and processes under varying cognitive load. Whilst demand is a broadly used term in the literature, it is limited by a lack of cross-domain operationalisation, and thus can be a confound in many inter-domain investigations. To address this, we generated and validated two large-scale, open-source, widely applicable stimuli sets that: (i) can probe a wide, continuous range of objective behaviourally-measured processing demand, and (ii) can be directly matched across two highly contrastive cognitive domains (semantic and visuospatial processing). We normed these tasks in a large, online cohort of healthy adults, enabling future studies to utilise and benefit from reliable baseline measures of demand. Using a matching algorithm that takes into account both speed and accuracy, we then closely paired each semantic judgement trial with a trial of comparable difficulty from the corresponding visuospatial task to produce a cross domain matched set of over 400 items. Finally, we analysed the psycholinguistic and perceptual properties of these stimuli to explore some of the underlying factors driving demand.

Humans must store and retrieve distinct implicit motor memories in different contexts. Recent research has revealed the motor adaptation system's sensitivity to contextual cues related to posture, spatial goals, and visual inputs. However, to date, the role of one critical variable in implicit motor memory retrieval - the passage of time - remains underexplored, even though it is arguably more fundamental. Here we ask if different temporal intervals can act as effective contextual cues for retrieving competing motor adaptation memories at the subsecond level. We combined a visuomotor dual adaptation paradigm with a 'set-go' task to link opposite perturbation contexts with different movement preparation intervals. The preparation intervals (and their associated perturbations) were randomly interleaved during training, leading to consistent interference. After training, a testing phase revealed robust contextual modulation effects, such that distinct implicit motor memories were retrieved based on the time elapsed during the pre-movement preparatory interval. These contextual memory effects generalized to novel preparatory intervals in a highly structured fashion. Contextual motor memory separation was seen both when elapsed time had to be internally tracked (high uncertainty) and when it was aided by visual and symbolic cues (low uncertainty). Learning and generalization data across experiments could be parsimoniously explained by a computational model that indexed motor memories via a cerebellar granule-cell-like temporal basis set that tiled the delay interval. Our findings suggest that intrinsic timing processes in the cerebellum may function to index motor memories and generalize those memories to new temporal contexts.

Musical training produces behavioral gains that extend beyond the practiced context. Yet, it is unclear whether such generalization is accompanied by broadly transferable neural changes or whether training-induced plasticity remains tied to the specific features of the learned material. In this study, 80 musically naive adults completed four piano-training sessions and underwent functional MRI before and after training. Twenty untrained participants served as passive controls. Participants performed a passive listening task in which they listened to the trained melody (Fur Elise) and an untrained melody (Ode to Joy), each played on a trained musical instrument (piano) and an untrained musical instrument (saxophone). Following training, participants demonstrated improved accuracy and rhythmic performance when playing an unfamiliar melody, indicating behavioral generalization. Training-induced neural changes, however, were predominantly associated with the learned melodic content. Listening to Fur Elise elicited increased activity in the intraparietal and supplementary motor cortex and in the cerebellum, as well as decreased activity in the superior temporal cortex. Consistently, pre- and post-training scans were most accurately classified based on brain activity to Fur Elise played on piano or saxophone, rather than brain response to Ode to Joy, suggesting that neural reorganization was driven more strongly by the practiced melody than by the trained instrumental timbre. Training also enhanced the neural separability of both melodic and instrumental information. Critically, pre-training responses to Fur Elise played on the piano predicted subsequent playing accuracy, whereas pre-training sensitivity to instrumental timbre predicted rhythmic performance in a novel melodic context. Together, these findings suggest that while acquired musical skills behaviorally transfer to novel material, practice-induced functional plasticity remains predominantly melody-specific. We conclude that musical skill acquisition reflects complementary contributions of pre-existing neural characteristics and learning-induced plasticity, with baseline auditory representations potentially supporting successful behavioral transfer.

The cerebellum has increasingly been implicated in Theory of Mind (ToM), a hallmark of human social cognition. Yet its role in the development of social understanding remains poorly understood, despite evidence linking early cerebellar disruptions to profound social cognitive deficits. Although explicit ToM reasoning emerges only around four years of age, preverbal infants already excel at predicting others actions, raising the question of how the cerebellum supports social cognition early in development. Here, we investigated structural cerebellar correlates of ToM (explicit false belief understanding) and nonverbal action prediction in children aged 3-4 years, a critical developmental period during which explicit ToM emerges. Greater gray matter volume in Crus II, a core node of the adult cerebellar ToM network, was associated with better explicit ToM performance. By contrast, nonverbal action prediction was linked to distinct, non-overlapping clusters in inferior lobule VIIB and the posterior vermis, regions implicated in action observation and salience processing in adults. These cerebellar regions further exhibited dissociable patterns of covariance with cerebral networks, linking Crus II to the canonical ToM network and VIIB/vermal regions to salience-related areas. Our findings reveal neuroanatomically distinct cerebellar substrates supporting two separable components of early social cognition: one anchored to explicit mental-state reasoning and another to an earlier-emerging nonverbal action prediction system potentially grounded in the processing of salient social cues. Together, these findings identify the cerebellum as a key contributor to multiple stages of social cognitive development and suggest that distinct cerebellar systems scaffold the emergence of mature social understanding.

Second language (L2) processing engages both language-specific and domain-general control systems, yet how these systems vary with L2 proficiency remains unclear. We used functional magnetic resonance imaging (fMRI) to examine neural activity during L2 English processing in Korean-English (K-E) bilinguals across three proficiency levels (beginner, intermediate, advanced). Participants performed rhyme and spelling judgement tasks manipulating orthographic-phonological conflict. Behaviourally, conflict conditions reduced accuracy, with proficiency effects observed selectively in the rhyme task. fMRI results showed that conflict processing recruited frontoparietal control regions, including inferior frontal and parietal cortices, accompanied by deactivation in default mode network regions. Critically, proficiency-related effects differed by task. During rhyme judgement, advanced bilinguals showed greater activation in the left supramarginal gyrus (SMG) and cerebellum, whereas intermediate bilinguals exhibited greater recruitment of the left middle orbital gyrus and dorsomedial prefrontal cortex. During spelling judgement, advanced bilinguals showed greater thalamic activation alongside greater deactivation of the right dorsolateral prefrontal cortex. Activity in the left SMG and cerebellum was positively associated with L2 reading score, and cerebellar activity was also associated with rhyme-task performance, whereas right DLPFC activity was negatively associated with the scores. These findings suggest that increasing L2 proficiency is associated less with altered recruitment of core reading regions than with task-specific shifts in the balance between phonological-specialized, subcortical attentional, and domain-general control systems supporting L2 processing.