Future self-continuity (FSC) and achievement motivation are two psychological constructs that facilitate future-oriented cognition and goal-directed behavior. However, the neural circuits integrating these constructs remain understudied from a cognitive neuroscience perspective-specifically how self-referential processing (a core component of FSC) interacts with reward valuation supported by the caudate. To address this question, we explored the relationship between future self-continuity and achievement motivation using a combination of resting-state functional connectivity (RSFC) and mediation analysis. Behavioral results confirmed a positive correlation between FSC and achievement motivation, while mediation analysis further revealed that left rACC-left caudate functional connectivity statistically accounted for part of this relationship. These findings suggest that rACC-caudate connectivity may reflect the integration of self-referential processing and reward valuation, providing a potential neural correlate for the interplay between future self-perception and achievement striving.
Adolescence is characterised by ongoing maturation of cortical circuits, yet little is known about how cortical responsiveness to repeated sensory stimulation changes across this developmental period. To address this question, we examined visual cortical responses to high-frequency sensory stimulation (HFS) in early adolescence, late adolescence, and adulthood using a sensory visual tetanisation paradigm. Visual evoked potentials (VEPs) were recorded from occipital and parieto-occipital regions during low-frequency visual stimulation (0.83 Hz) at baseline and 2, 4, and 20 min following HFS (8.87 Hz, 2 min). Post-HFS changes in the N1b and P2a VEP components, indexing early and later stages of visual processing, respectively, were used to quantify changes in cortical responsiveness. Across all age groups, N1b amplitudes showed sustained increases in both occipital and parieto-occipital regions following HFS, with no significant effects of age. In contrast, P2a amplitudes demonstrated age- and region-specific modulation. At occipital sites, all groups exhibited a comparable transient attenuation of P2a following HFS that returned to baseline levels. However, in parieto-occipital regions, P2a attenuation persisted across all post-HFS assessments in early adolescents, whereas in late adolescents and adults it recovered to baseline by 20 min post-HFS. These findings reveal a developmental dissociation between early and later stages of visual processing, with younger adolescents showing more sustained modulation of later neural responses. Together, the results suggest continued maturation of the mechanisms regulating cortical responsiveness to sensory stimulation across adolescence.
Individuals with Internet Gaming Disorder (IGD) exhibit heightened engagement with gaming alongside diminished responsiveness to natural rewards such as food. Whether this dissociation extends to reward learning, a core process linking outcomes to future behavior, remains unclear. Using a Probabilistic Reward Task combined with event-related potentials, we examined reinforcement and extinction learning across food and gaming rewards in individuals with IGD (n = 50) and recreational game users (RGUs; n = 50). In the reinforcement learning task, the IGD group showed significantly smaller response biases and attenuated Reward Positivity (RewP) amplitudes compared to RGUs when processing food rewards, whereas no significant group differences emerged for gaming rewards. In the extinction learning task, the IGD group exhibited smaller response biases and reduced RewP amplitudes across both reward types compared to RGUs, with the most pronounced impairment for gaming punishments. Additionally, the IGD group displayed larger Late Positive Potential (LPP) amplitudes for gaming reward versus neutral feedback, a pattern not observed for food rewards. Across both experiments, RewP amplitude was positively correlated with response bias and negatively correlated with IGD symptom severity. These findings provide converging behavioral and neurophysiological evidence for domain-specific deficits in reward learning in IGD, characterized by impaired reinforcement from natural rewards and deficient extinction from gaming punishments. Selective dysfunction in reward learning may represent one neurocognitive pathway through which gaming behaviors are strengthened and maintained at the expense of natural reward processing in IGD.
The neurovisceral integration model posits that the reciprocal interaction between the amygdala and prefrontal cortex (PFC) influences heart rate variability (HRV) and emotion modulation. Emerging evidence also suggests that HRV influences brain regions involved in emotion regulation. Building on previous functional magnetic resonance (fMRI) studies, the present study employed structural covariance analysis to examine the effects of HRV biofeedback intervention on the structural covariance between the amygdala and cortical regions. Investigating the covariation between the amygdala and cortical regions during a biofeedback intervention can provide insight into how HRV regulatory attempts influence neural structures involved in HRV and emotion regulation. Baseline and post-intervention MRI scans were collected from 151 participants (100 younger and 51 older adults) who completed 5 weeks of daily biofeedback sessions to either increase (Osc + condition) or decrease (Osc- condition) HRV. Across both age groups, the two intervention conditions were associated with opposing patterns of structural covariance between the right amygdala and the left medial prefrontal cortex (mPFC). In the Osc + condition, volume changes in the left amygdala showed an inverse association with volume changes in the right orbitofrontal cortex (OFC), while no association was observed in the Osc- condition. These findings are consistent with the neurovisceral integration model perspective that the PFC exerts inhibitory influence over subcortical structures and that HRV influences brain regions involved in emotion regulation.
Growing evidence links amyloid accumulation to altered gamma activity, positioning gamma oscillations as a promising biomarker for Alzheimer's pathology. We examined resting state gamma activity in early-onset Alzheimer's disease (EOAD) patients confirmed by CSF biomarkers. We aim to provide a holistic view of gamma activity patterns in EOAD by integrating structural and functional neuroimaging. In this cross-sectional study, we compared gamma power and coherence-total gamma (30-48 Hz), and sub-bands (gamma-1: 30-35 Hz; gamma-2: 35-40 Hz; gamma-3: 40-48 Hz)- alongside gray matter volumes in EOAD patients (N = 24) and matched healthy controls (N = 24). We further assessed gamma-related changes in relation to CSF biomarkers, cognitive performance, and structural brain alterations. Finally, discriminant analysis evaluated classification performance of these measures. EOAD patients showed increased gamma power, with region-specific increases and decreases in coherence. Frontal gray matter volume negatively correlated with coherence, whereas posterior volumes showed positive correlations. A combined model of parietal gamma power and coherence achieved 83.3% classification accuracy, with a cross-validated accuracy of 77.1% (ROC-AUC = 0.844). Overall, we showed that the gamma activity does not follow a strict pattern in EOAD pathology. Gamma power is elevated, whereas connectivity patterns vary by brain region, closely mirroring structural brain alterations. These findings underscore the relevance of gamma activity in Alzheimer's research and open new avenues for future investigation. CLINICAL TRIAL REGISTRATION NUMBER: NCT05989087. NAME OF THE TRIAL REGISTRY: The Resting-state EEG Gamma Oscillations in Alzheimer's Disease.
We spend approximately one-third of our lives asleep. This behavioral state plays a pivotal role in protecting physical and mental health. A lack of sleep has negative consequences on the brain and behavior. Electrophysiological studies have shown that gamma oscillations are involved in sleep-wake processes. However, gamma peak frequency and power characteristics remain unclear during a night of sleep deprivation. Forty-four (44) healthy young adults were enrolled in this EEG study. Participants were randomly assigned to a rested control group (RC: n = 22, 13 female) or a night of sleep deprivation group (SD; n = 22, 12 female). The SD group was required to stay awake 24 h before the morning EEG session, whereas the RC group maintained their 7-9 h of habitual night sleep before the session. Sleepiness evaluations were conducted before the EEG for both groups. Eight minutes of resting-state EEG were recorded for all participants (4 min. eyes-open, 4 min. eyes-closed). Peak frequency (Hz) and power (μV2) values for gamma frequency band (28-48 Hz) were investigated. Our results revealed that one night of SD group rated their sleepiness significantly higher than the RCs. In EEG results, gamma peak frequency of the SD group in the eyes-open condition was significantly lower than that of the RCs. In contrast, power values were not different between groups. In light of our findings, the neural mechanisms of gamma oscillations and their potential links to sleep deprivation are discussed.
Developmental dyscalculia (DD) may involve difficulty using prior context during cognitive control. We tested whether children with DD showed reduced cue-context modulation of theta recorded at Fz and Cz. Single-trial electroencephalographic data were analyzed from 143 children aged 6-11 years, including 71 with DD and 72 typically developing (TD) children. The task was a Cue/Uncue Go/No-Go paradigm. The primary contrast compared S6 and S8, two no-response conditions with the same current stimulus category but different preceding cue contexts. To address the original montage asymmetry, all revised analyses used the same 19-electrode montage and reference. The primary EEG sample included 128 children with identifiable Fz and Cz channels and at least 20 artifact-free epochs in each condition. Three-cycle Morlet wavelets quantified the S6-minus-S8 log-power contrast at 5-7 Hz and 200-400 ms. Total-theta modulation was smaller in DD than TD, F(1,124) = 15.86, p < .001, partial η2 = 0.113. The effect remained after removal of condition-averaged event-related potentials, F(1,124) = 11.66, p < .001, partial η2 = 0.086. It also remained after S8 trial-count matching and joint adjustment for technical variables and task performance. Greater modulation showed modest associations with more hits and faster responses. These findings indicate reduced S6-minus-S8 theta modulation at Fz/Cz in this DD cohort. The contrast is not specific to arithmetic, DD, or a single control process, and it should not be interpreted as a source-localized or individual diagnostic marker.
Disgust is often conceptualised as a basic emotion arising from contamination avoidance; however, the functional expansion account distinguishes multiple disgust domains that differ in their social meanings and behavioural outcomes. This study investigated whether core disgust, animal-reminder disgust, and moral disgust exhibit distinct whole-brain temporal dynamics indexed by electroencephalography (EEG) microstates. Forty university students viewed seven video clips eliciting the three disgust types and neutral affect while their EEG activity was recorded. They also completed tests of emotional state and disgust sensitivity. EEG microstate analysis was used to identify four canonical microstate classes and quantify their temporal and transition indices across conditions. An exploratory machine-learning analysis using L1-regularised logistic regression was also conducted to identify candidate microstate features differentiating disgust from neutral conditions. The mean duration of microstate D made the strongest positive contribution to disgust classification. Subsequent linear mixed-effects models indicated an uncorrected tendency towards longer microstate D duration for moral disgust than that for neutral affect, although this effect was deemed insignificant after FDR correction. Exploratory comparisons among disgust conditions further suggested longer microstate D duration during moral and animal-reminder disgust than that observed for core disgust; however, the corrected main effect of disgust type was insignificant.These findings suggest that microstate D duration may reflect disgust-related temporal dynamics, particularly for stimuli requiring more elaborate appraisals, such as moral and animal-reminder disgust. Although the observed effects were exploratory and not robust after correction for multiple comparisons, we demonstrate the potential utility of EEG microstate analysis for examining the temporal organisation of complex disgust processing.
The present short-term longitudinal study examined the associations between preschoolers' coping responses to their personal experience of peer victimization and later internalizing symptoms. This study prospectively tested autonomic nervous system reactivity (i.e., respiratory sinus arrhythmia reactivity, RSA-R; and skin conductance level reactivity, SCL-R) as moderators of these associations. This secondary data analysis assessed a relatively well-resourced sample of 83 preschoolers (M = 45.59 months old, SD = 3.80, 47.0% female) over the course of approximately 1 year: first during the spring of participants' preschool (Time 1) and again during the spring of the subsequent pre-kindergarten year (Time 2). Hypotheses were guided by polyvagal theory, which highlights the importance of interactive effects of the sympathetic and parasympathetic nervous systems in the stress response. Results demonstrated that approach coping was associated with decreases in later internalizing symptoms. In addition, avoidance coping was associated with decreases in internalizing symptoms for participants exhibiting SCL-R inhibition. RSA-R did not moderate associations between coping and later internalizing symptoms. These findings provide preliminary support for the hypothesis that the adaptive alignment between coping behaviors and physiological reactivity may serve as a protective factor in buffering against psychological distress.
Cognitive reappraisal reliably reduces subjective negative affect and attenuates the late positive potential (LPP), a well-established neural marker of sustained motivational relevance and affective salience. However, it remains unclear whether peripheral cardiac responses reflect regulatory success in a manner comparable to cortical indices. The present study examined whether evoked cardiac responses (ECRs) are modulated by reappraisal and whether such modulation converges with changes in the LPP. Fifty-two young adult women viewed neutral and negative images and were instructed either to passively view or to reappraise negative stimuli. Subjective affect, LPP amplitude, and stimulus-locked cardiac deceleration were assessed on each trial. Reappraisal reduced reported negative affect and attenuated the LPP relative to passive viewing of negative stimuli, replicating established findings. In contrast, reappraisal enhanced cardiac deceleration rather than attenuating it. To assess regulatory success across measures, we computed difference scores reflecting reappraisal-related changes (reappraisal minus passive viewing of negative stimuli) for subjective affect, the LPP, and cardiac responses. Behavioral reappraisal success was modestly associated with both neural and cardiac difference scores, whereas the neural and cardiac indices were not significantly related to each other. These findings suggest that neural and cardiac responses during reappraisal reflect partially dissociable processes.
Whether arithmetic and language rely on shared or distinct processing mechanisms remains unresolved, particularly at the level of semantic and structural anomaly processing. The present EEG study addressed this question by comparing three written expression types: symbolic arithmetic, verbal arithmetic in Russian, and factual statements in Russian. Thirty-nine native Russian-speaking adults read segmented expressions and made explicit judgments about semantic correctness, while structural violations remained task-irrelevant. Semantic and structural correctness were manipulated orthogonally, and event-related potentials were analyzed with a priori focus on the N400, LAN, and P600. Semantic anomalies elicited a robust N400-like response across all three expression types, indicating broadly comparable meaning-related anomaly processing. Structural effects, however, were more dependent on representational format and processing stage. In earlier time windows, the clearest 300-500 ms structural effect was observed in symbolic arithmetic under the present local anomaly manipulations, although this early difference may partly reflect greater transparency of the symbolic structural violation. In the later P600 range, both symbolic arithmetic and factual statements showed significant structural-anomaly positivities, whereas no reliable comparable effect was detected for verbal arithmetic. These findings argue against both a fully shared and a fully segregated account of arithmetic- and language-related processing. Instead, they suggest a stage-dependent pattern of partial convergence and divergence. Semantic anomaly processing was broadly similar across expression types, whereas structural anomaly processing was more format-sensitive. Verbal arithmetic did not show a stable intermediate ERP profile across components under the present task and materials, suggesting that verbalized arithmetic should not be treated simply as symbolic arithmetic in words or as ordinary language with numerical content.
The visual world is complex; however, we still access its richness despite computational limits in our visual system. Ensemble perception overcomes this bottleneck by extracting statistics, such as the mean, standard deviation, or other distributional characteristics, of feature sets like orientations or facial expressions, without processing each object individually. The role of attention in ensemble processing is debated: can we extract statistical information about a set when we do not attend to it? To address this question, we used the Posner cueing paradigm in which participants (N = 30, Mean Age = 19.6) reported the orientation of a single Gabor patch or the mean orientation of an ensemble of Gabors. Ensembles consisted of several Gabors whose orientations were sampled from a truncated Gaussian distribution centered on a clockwise or counterclockwise mean. A central arrow pre-cue pointed to either the right or left prior to bilateral stimulus presentation, and a post-cue indicated which side should be reported. In 20% of trials, the post-cued location differed from the pre-cued, so attention was misdirected (invalid trials); in others, cues aligned (valid trials). We used EEG to confirm attentional deployment to the cued location, and we used behavioral measures to estimate the effect of such attentional orientation. We looked for the N2pc, an EEG marker of lateralized attention, but could only detect it in the single Gabor condition. However, multivariate pattern analysis (MVPA) confirmed attentional shifts by showing above-chance classification accuracy of the attended side in both single and ensemble conditions. Behavioral performance was measured using d', a measure of sensitivity. Performance was impaired on invalid compared to valid trials in both the single Gabor and ensemble conditions, consistent with a role of attention in both cases. Accuracy dropped to chance levels on invalid ensemble trials but remained above chance, though reduced, on single trials. Overall, these findings indicate that misdirecting spatial attention severely impairs the ability to report the mean orientation of peripherally presented ensembles.
Previous neuroimaging studies indicate that the inferior frontal gyrus (IFG) may be associated with emotion regulation through the use of cognitive reappraisal strategy. However, whether there is a causal role of IFG in reappraisal-based regulation of general negative emotions remains unclear. Therefore, the present study employed a two-study, progressive research framework combining functional magnetic resonance imaging (fMRI) and transcranial direct current stimulation (tDCS) to investigate the critical role of the IFG in reappraisal-based emotion regulation. In Study 1 (fMRI experiment), thirty-two participants completed an emotion regulation task during scanning. Brain activations were compared between reappraisal condition and view condition to identify neural correlates of emotion regulation. Whole-brain family-wise error (FWE p < 0.05) correction revealed significant activation in the bilateral IFG and other frontal-temporal regions during the use of reappraisal strategy. In Study 2 (tDCS experiment), building on the fMRI findings, twenty participants were recruited to complete the emotion regulation task while receiving counterbalanced active or sham anodal tDCS over the right IFG (one week apart), to observe potential changes in regulation effect when using the reappraisal strategy. Results showed that active tDCS significantly enhanced the regulation effect of reappraisal compared to sham stimulation. Collectively, these findings provide converging evidence for a critical role of the right IFG in reappraisal-based down-regulation of general negative emotions. This may have potential implications for clinical interventions, particularly in psychiatric conditions associated with emotion regulation deficits.
Slow-paced breathing (SPB) interventions have shown promise for modulating physiological regulation and enhancing cognitive performance. However, few studies have empirically examined how specific components of the breathing structure, particularly inhalation-holds during SPB, influence autonomic and endocrine responses. Thus, the present pilot study sought to address this gap by comparing the physiological and psychological outcomes of SPB with and without a 4-s post-inhalation breath hold using a within-subjects design. Participants (n = 32; 37% female) completed two SPB conditions in a randomized order: a prolonged-exhalation pattern (4-6) without a breath hold and a matched pattern with an inhalation-hold (4-4-6). Measures of hypothalamic-pituitary-adrenal (HPA) axis activity (salivary cortisol), cardiac autonomic regulation (RMSSD), sympathetic activation (skin conductance), and self-reported stress, arousal, and mood were collected. Results indicated that the inhalation-hold condition produced significantly greater decreases in cortisol and RMSSD than the no-hold SPB condition, although supplementary HF power analyses did not replicate the RMSSD interaction effect. Measures of tonic skin conductance increased over time in both conditions, and no differences between conditions were observed. Neither breathing pattern significantly impacted self-reported stress, arousal, or mood. These findings suggest that a four-second inhalation-hold can modulate HPA activity during SPB with a prolonged exhalation, but the respiratory mechanics of the hold may alter short-term respiratory-cardiac dynamics reflected in RMSSD.
Trait anxiety is a clinically relevant vulnerability dimension, yet resting-state EEG work has relied predominantly on stationary spectral summaries. We tested whether resting EEG microstate transition dynamics index trait-anxiety-related variation more consistently than conventional alpha (8-13 Hz)- and theta (4-8 Hz)-band power. Self-report measures and approximately 6 min of eyes-closed resting EEG were obtained from 333 undergraduates. HC3-robust regressions contrasted three prespecified microstate transition metrics with two prespecified spectral indices. Follow-up analyses comprised family-wise transition tests, supplemental spectral summaries, repeated 5-fold cross-validated model comparisons, bootstrap resampling, extreme-groups sensitivity analyses, and exploratory Gaussian mixture modeling. Trait anxiety was positively related to state anxiety, anxious arousal, intolerance of uncertainty, maladaptive metacognitive beliefs, and depressive symptoms. Higher trait anxiety was associated with lower deviation-from-chance B → A switching, lower observed B → A transition probability, and higher deviation-from-chance B → C switching. After adjustment for depressive symptoms, intolerance of uncertainty, and metacognitive beliefs, these microstate associations remained significant, whereas frontal alpha and midline theta did not remain significant after this psychosocial adjustment. In direct model comparisons, the base + microstate model outperformed the base + spectral model, and bootstrap resampling showed highly stable coefficient direction for the three prespecified microstate metrics. In this young-adult sample, trait anxiety was more consistently associated with the temporal organization of resting brain states than with stationary spectral power.
Emotion recognition ability (ERA) - the capacity to accurately infer emotions from nonverbal facial, vocal, and bodily expressions - is a fundamental socio-emotional skill linked to improved social functioning, clinical counseling efficacy, workplace performance, and interpersonal relationships. While numerous ERA training programs for healthy adults have demonstrated improvements in ERA as measured through standardized tests, evidence of transfer effects to real-world socio-emotional processes remains limited. This study investigates whether a short multimodal ERA training (Training of Emotion Recognition Ability, TERA) impacts visual attention to emotional stimuli and emotional reactivity compared to a control attention and memory training in a randomized controlled trial (RCT). Participants (planned sample of N = 230) will be randomly assigned to either the TERA or control group. Visual attention will be assessed using eye-tracking while participants view videos of autobiographical emotional accounts. Emotional reactivity will be measured through self-reported emotional congruence (alignment of participants' emotions with those of video protagonists) and sympathy for protagonists. We expect that the TERA group will exhibit longer dwell time on protagonists' eye regions, greater emotional congruence and sympathy, and more accurate judgments of protagonists' displayed emotions compared to controls. Additionally, mediation analyses will investigate whether heightened attention to the eyes explains the training's effects on emotion recognition accuracy, emotional congruence, and sympathy. This study addresses critical gaps in the literature by examining the transfer effects of ERA training on attentional and emotional processes using ecologically valid stimuli. Findings could inform the development of targeted interventions to enhance socio-emotional functioning in diverse populations.
Previous studies have found that emotional words as a second visual target (T2) are detected more accurately than neutral T2 in a Rapid Serial Visual Presentation (RSVP) task, a phenomenon known as the antagonistic effect of the Attentional Blink (AB). However, it remains unclear whether the antagonistic effect of AB for emotional words is modulated by T1 task difficulty. In this study, we examined whether T1 difficulty moderates the AB effect for T2 (emotional words: negative, positive, and neutral) in RSVP using event-related potentials (ERPs). Participants were instructed to perform a letter search task for T1 and an emotional classification task for T2. Behavioral results showed that the antagonistic effect of AB for emotional words as T2 was not moderated by T1 difficulty. However, ERP results revealed that negative words elicited larger P3a and P3b amplitudes compared with neutral words, and positive words elicited larger P3b amplitudes over neutral words when T1 difficulty was low, whereas no significant differences were observed among the three emotional word types when T1 difficulty was high. Overall, we conclude that attentional resources constrain the neural expression of the antagonistic effect of the attentional blink for emotional words.
The relationship between attentional resource allocation and early motor skill learning remains unclear. This study examined how proficiency in drone operation influences attention and which factors contribute to individual differences. Twenty-three participants with no prior drone experience performed a simulator task consisting of 12 three-minute trials, divided into three phases of four trials each. We estimated attentional resource allocation using auditory-evoked potentials (AEPs) elicited by task-irrelevant auditory probes. Performance improved significantly across phases (ps < 0.001), and N1 and P2 amplitudes were significantly reduced (ps < 0.05), indicating that participants allocated more attentional resources to drone operation as proficiency increased. Furthermore, overall task performance significantly correlated with smaller P2 amplitudes (r = -0.42, p = .044). Given that variability in P2 amplitude of AEPs elicited by task-irrelevant probes reflects endogenous attention, the findings suggest that endogenous attention might play a key role in early motor skill learning and underlie individual differences.
Developmental dyscalculia (DD) is associated with persistent arithmetic difficulties, but the temporal dynamics of operation-specific arithmetic processing remain unclear. This study examined operation-locked event-related potentials (ERPs) during addition, subtraction, and multiplication in 76 children aged 8-12 years, including 37 children with DD and 39 typically developing (TD) controls. ERPs were time-locked to the onset of the second operand, when the complete arithmetic expression became available. Children with DD showed lower accuracy and longer reaction times across all operations. ERP analyses revealed prolonged N1, P2, and N2a latencies, together with reduced N1 and N2a amplitudes and increased frontal P2 amplitude. Subtraction produced the most evident delay, particularly in early latency measures. Across the full sample, longer N2a latencies were associated with slower reaction times, although within-group correlations were not significant. These findings indicate that DD is characterized by a temporal processing inefficiency that begins during early symbolic encoding and extends to intermediate monitoring and updating of arithmetic information, especially when procedural computation demands are high.
Humans often rely on social group information to anticipate others' actions. While prior research suggests that group members' goals can be generalized to new individuals, it remains unclear whether this process scales incrementally with observed evidence or is constrained by prior beliefs. Across two experiments, we combined behavioral and EEG measures to examine whether the prevalence of observed actions enhances goal-directed actions generalization. In Experiment 1, participants observed animated groups performing goal-directed actions toward different targets while EEG was recorded. Both behavioral facilitation effects and contingent negative variation (CNV) amplitudes revealed stronger goal-based expectations for a group member when the group pursued a shared goal (i.e., high prevalence); medium and low prevalence conditions did not differ. Such a nongraded pattern indicates that statistical evidence is not the only factor shaping goal-based action generalization. In contrast, Experiment 2 showed that movement-based action generalization increases in a graded manner with action prevalence, highlighting the specific deviation from a graded pattern in goal-based action generalization. Together, these findings provide convergent behavioral and neural evidence that goal-based action generalization cannot be explained entirely by statistical evidence, but is instead constrained by shared-goal beliefs, underscoring the role of group-related beliefs in action prediction.
