Communications Psychology

Confidence evaluates the likely accuracy of a current decision. However, to be maximally informative about accuracy, confidence judgments should incorporate information about ones broader decision tendencies, such as their propensity to favor specific alternatives. We distinguish bias-aware confidence, which considers such tendencies, from bias-blind confidence, which relies only on evidence available on the current trial. To adjudicate between bias-aware and bias-blind confidence, we identified a signature of bias-aware confidence: the down-weighting of confidence for alternatives that a participant is biased toward. We then used a large dataset (N = 200) spanning 4- and 8-choice digit-classification tasks to show that humans reliably exhibit this signature of bias-aware confidence. This effect was reduced under speed pressure and could not be explained by guessing. In contrast to the human results, artificial neural networks (ANNs) trained for object recognition lacked this signature of bias- aware confidence. Importantly, augmenting ANNs with a metacognitive module that allows confidence to take the networks biases into account led to the emergence of human-like bias- aware confidence. These findings show that human confidence incorporates not only information from the current trial but also longer-term decision tendencies, and that this capacity - absent in standard ANNs - can be conferred through specialized metacognitive mechanisms.

Why do multi-attribute choices so often depart from classical weighted-additive decision rules? Rather than attributing such deviations solely to biases or heuristics, we propose a resource-rational account in which value differences are encoded via capacity-limited information channels. Under resource-rational compression, these difference representations are systematically distorted, such that behavior deviates from weighted-additive predictions because value differences are not represented veridically. This theoretical account makes testable predictions about power-law relationships between true and internally represented differences. The amount of power-law-like compression is determined by information-processing capacity, emergent long-tailed prior distributions over attribute differences, and, in choice contexts, goal-dependent subjective weights that govern the allocation of limited capacity across attribute channels. We test and find support for these predictions in an attribute difference-estimation task and by reanalyzing existing food- and social-choice datasets. These results provide converging evidence for a normative, information-theoretic account of systematic nonlinearities in multi-attribute decision making. Together they show how goals can interact with cognitive capacity and priors to shape representational precision in ways that may facilitate or impair decision making.

Human perception is continuously shaped by internal cognitive states, yet how motivation and working memory jointly influence perceptual sensitivity remains poorly understood. Here, we combined behavioural experiments, computational modelling, and pupillometry to determine whether motivation enhances perception by amplifying working memory (WM)-driven facilitation or whether both exert independent influences. Participants performed a near-threshold visuospatial discrimination task under systematically manipulated motivational and WM states. Behaviourally, both motivation and WM independently improved perceptual performance, producing the greatest enhancement when both were present. A Bayesian generalized linear psychometric model revealed that these improvements were best explained by independent additive contributions to effective perceptual sensitivity, rather than motivational amplification of WM. Consistent with this computational framework, pupil dynamics tracked trial-by-trial fluctuations in effective perceptual sensitivity while revealing temporally dissociable influences of WM and motivation during perceptual decision making. Together, our findings demonstrate that the joint effects of motivation and working memory arise from distinct computational and temporal mechanisms, providing a unified framework for understanding top-down regulation of human perception.

Self-supervised foundation models of aging are increasingly built from longitudinal data (biobanks, electronic health records, wearables) that is inherently incomplete: no individual is followed across a whole lifetime, and how much of each life is captured varies widely. This raises two linked questions: is it worth modeling an individual's whole life course rather than its current state, and can such a model be built from brief, fragmentary records? No human cohort can settle them, because none offers a complete life to compare against. We turn to the African turquoise killifish (Nothobranchius furzeri), tracked from youth to natural death in publicly released recordings, as a controlled testbed: its complete lifespans provide the full-life reference that human data lacks. On these data we build LifeMAE, a two-stage selfsupervised model: a day encoder that summarizes each day of behavior, then a life-course encoder over the trajectory of those daily summaries. We find that the day encoder alone is already strong: from a single day of behavior it predicts chronological age, separates long- from short-lived individuals (coarsely), and flags nearness to death. Adding the life-course encoder improves on none of the three; each is matched by trivially aggregating the day-level predictions (a smoother for age, an early-life average for lifespan). Near-term mortality seems the exception, where the whole-life model looks far better (AUROC 0.81 to 0.91), but the gain is not behavioral: it reflects where each day falls within the observation window (a cue supplied by the model's encoding of time), and a single-day model given that cue closes the gap at any observation length. For these traits, an individual's place in its life course is legible from a single day: the trajectory stage is unnecessary, and the record it needs is as short as one day, the finest grain our day-level setup resolves. For characterizing a cohort, this favors observing many individuals briefly over tracking a few for long. The result joins a growing body of work in which deep and foundation models, fairly benchmarked, fail to beat deliberately simple baselines. We add a concrete mechanism for the over-optimism: a model's encoding of time can leak the very quantity it predicts, which backwardlooking evaluation mistakes for learned biology, so only evaluation fixed to the moment of prediction is trustworthy.

Humans exhibit a pervasive drive toward self-consistency, often failing to revise previous decisions even when confronted with contradictory evidence. Here, we investigate the computational mechanisms underlying decision revision in perceptual tasks, examining the regulatory role of metacognition. To do so, we capitalize on a novel paradigm in which participants are repeatedly presented with identical sensory information and allowed to revise their choices after each exposure. Our results reveal that repeated exposure to the same stimulus systematically biases subsequent judgments toward prior responses. Using drift-diffusion modeling, we tested competing explanations incorporating different assumptions about how prior choices affect evidence accumulation. Our findings indicate that consistency biases emerge from asymmetric sensory weighting, selectively amplifying information consistent with previous choices--a phenomenon akin to confirmation bias. Crucially, individuals with higher metacognitive skills exhibited weaker confirmatory biases and more flexible integration of repeated sensory information, enabling greater adaptability in decision-making. These findings highlight the continuous nature of perceptual inference and underscore metacognitions pivotal role in mitigating bias and optimizing decision flexibility.

Prediction in dynamic situations, in which relevant elements evolve over time, is a fundamental cognitive function. The brain relies on specialized predictive mechanisms, including time compaction, a process that supports dynamic processing by embedding temporal information into space and transforming future interactions into salient spatial representations. Here we investigated how future interactions are salient during dynamic events and how this salience shapes behavior. Participants performed a visuomotor prediction task in which they estimated the future trajectory of a moving object after observing only the initial portion of its motion, while another object was simultaneously present and could generate either interactive (collision) or non-interactive (crossing) dynamics. Although accurate performance required extrapolating motion solely from kinematic information, participants predictions were systematically biased toward locations associated with future interactions. Prediction accuracy was reduced in situations involving potential future interactions compared to non-interactive dynamics. Importantly, participants consistently responded closer to predicted interaction points, even when this strategy did not improve accuracy or trajectory extrapolation. Substantial inter-individual variability was observed, revealing conservative and risk-taking predictive strategies with systematic group differences. When participants were explicitly instructed to improve performance, overall accuracy improved only marginally, while predictive behavior shifted toward greater reliance on interaction-related locations, particularly among those who had not already adopted this strategy. We propose that this interaction-driven bias reflects a core property of time compaction, supporting the idea that predictive cognition relies on future interactions as stable reference points under dynamic uncertainty.

Learning is typically understood as a process driven by prediction errors, when outcomes differ from expectations. Yet it remains unclear whether outcomes that perfectly match expectations are psychologically and computationally meaningful. Here, we tested whether zero prediction errors shape affect, belief updating, and neural feedback processing in human reinforcement learning. Participants repeatedly predicted rewards in environments varying in uncertainty, with a subset of trial outcomes manipulated to exactly match their predictions. Zero prediction errors produced the highest momentary happiness, and computational modeling showed that behavior was best explained by a model in which zero prediction errors induce a distinct latent belief state that guides subsequent updating, particularly under higher uncertainty and in individuals with greater intolerance of uncertainty. Outcome-locked EEG analyses further showed that zero prediction errors elicited distinct P3-like responses, with residual neural activity predicting attenuated updating after zero prediction errors but enhanced updating after standard prediction errors. These findings suggest that perfect predictions are not neutral, but informative events that actively shape affect, behavior, and neural feedback processing.

The ability to mentally rotate objects is a fundamental feature of human cognition, and humans can use this ability to make choices about objects based on their geometry. However, remarkably little is known about how such choices are reached, and what sort of visual information might facilitate them. We devised an experiment where participants had to mentally simulate an object's rotation to choose which of two objects was better for a subsequent task based on its shape alone. We also tracked their gaze while they made their choice, to see which visual information they were using to facilitate this mental simulation. We found that participants were consistently able to choose the most suitable object for the task, and, remarkably, the visual information they sampled was directly linked to their choices. Put simply, participants made better choices when they looked at more informative regions of the objects, and participants who sampled regions that were better for facilitating mental simulation made better choices overall. These findings reveal a direct link between fixations, simulation, and decision-making, suggesting that to perform any fine-grained mental simulation people need to direct their gaze at specific, informative points of an object to simulate its two-dimensional proximal image displacement.

Healthy older adults are more susceptible to false memories than young adults. Traditional false memory paradigms leverage semantic overlap, shared meaning, to induce false memories, but experiences can also overlap temporally whereby they occur close together in time. Prior work shows that older adults have impaired episodic memory, memory for events within a spatiotemporal context, corresponding to an overall shift toward semantic memory and away from episodic memory across the lifespan. We hypothesize that compared to young adults, older adults rely more heavily on semantic versus episodic information, which promotes false memory. We collected behavioral data in young and older adults performing an old/new recognition memory task in which we manipulated the degree of semantic and temporal overlap between study words and included critical lures, unstudied words that semantically overlap with study words. We find that whereas young and older adults are similarly reliant on semantic relative to episodic information to support false memory, the two age groups differ in their reliance on semantic relative to episodic information to support true memory. These results suggest that differences in the orientation of attention -- toward semantic vs. episodic information -- may underlie age-related memory changes.

Social decision-making depends on inferring others hidden preferences from observable behavior. Yet it remains unclear how humans combine choices with process cues such as response times and gaze when learning about others in real-time interaction. Here we combine a novel multi-attribute bargaining task with eye-tracking and show that multiple decision-process cues support preference inference. Across 75 buyer-seller dyads, buyers acceptance rates tracked offer utility, rejection speed reflected decision confidence, and first fixations preferentially targeted the highest-weighted attribute. Sellers adapted subsequent offers using choices, response times, and, when available, gaze cues. A hierarchical inference and choice model suggested that sellers balanced expected utility with expected information gain and updated their beliefs in a Bayesian manner. Although gaze access did not improve overall performance, it changed how sellers used attentional information. These findings shed light on how humans infer others hidden preferences from decision dynamics in real-time social interaction.

Flexible decision making depends on retrieving and recombining memories. Yet because this process unfolds covertly, its governing principles remain unknown. Here we use gaze reinstatement to uncover the hidden dynamics and computational logic of memory retrieval during flexible behavior. As people deliberated on a blank screen, they directed their gaze toward the encoding locations of decision-relevant experiences, and these fixations shaped their evolving choice. A task-optimized recurrent neural network captured both their behavior and gaze patterns by learning to balance retrieval costs against expected gains in decision quality. These results demonstrate that flexible decisions emerge from a resource-rational process in which memories are sampled to construct decision variables on the fly.

Human decision-making behavior varies widely across individuals and task conditions. This variability is often interpreted in terms of different suboptimal decision strategies, but the principles that govern these suboptimalities remain poorly understood. We propose that some of these suboptimalities can be understood in terms of limited-capacity, but information-efficient, inference processes that inform decision-making. We developed and used new theoretical and empirical approaches to compare the amount of information used (capacity) to the effectiveness with which it was used (accuracy) by individual participants performing simple inference tasks. Variable, suboptimal performance was explained largely by inference that had variable, limited information capacity. Across these capacity limits, and regardless of whether the inference strategy was based on optimal or heuristic principles, the information was used effectively to maximize accuracy for a given capacity. This form of flexible and efficient information bottleneck reflects fundamental capacity-accuracy tradeoffs that structure individual variability.

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.

INTRODUCTIONDo continuous cognitive totals capture patient-relevant transitions, or does decline have discrete structure? METHODSWe formalize a discrete cognitive resolution (DCR) model in which decline is loss of binary discriminative coordinates, with deterministic emissions tying items to a shared low-dimensional mask. Using natively discrete item-level data (ADAS-Cog, 688 ADNI participants; MoCA sub-items, 13,323 NACC participants), we tested pre-specified signatures by out-of-sample log-loss against continuous-drift, per-item, and mixed-effects IRT competitors, with a coordinate-label permutation null (S2). RESULTSDCR beat both pre-specified baselines (ADNI 0.436 vs 0.965; NACC 0.566 vs 0.602). S2 was decisive in ADNI (AUC 0.782; null 0.529, P < .001); in NACC the signal concentrated in orientation (AUC 0.718). Mixed-effects IRT achieved lower log-loss than DCR. DISCUSSIONCognitive decline shows discrete coordinate structure when items are single-coordinate probes. The claim is structural, not predictive; encoding is decisive.

Action selection involves two systems: a model-free reinforcement learning strategy, which relies on experience with action-outcome pairs, and a model-based reinforcement learning strategy, which enables more flexible behavior via inference using a model of the invariant environmental structure. Although environmental change requires more flexible behavior, the ability of volatility, a higher-order statistic that captures how rapidly or frequently the environment changes, to systematically modulate these strategies remains unclear. We examined the effects of reward volatility on arbitration between model-free and model-based reinforcement learning strategies using two modified two-step decision tasks. In Experiment 1, participants performed tasks with different levels of reward volatility and time pressure. In Experiment 2, we systematically manipulated reward volatility across a broader range to assess the relationship between volatility and learning strategy. Behavioral data were analyzed using model-agnostic one-trial and multitrial back analyses, reinforcement learning simulations, and hierarchical Bayesian model fitting. Across experiments, reward volatility exerted an inverse U-shaped nonlinear effect on the arbitration between model-free and model-based reinforcement learning strategies, as the model-based learning strategy was strongly driven at intermediate levels of reward volatility. These modulation effects were observed only in individuals who had learned the transition structure in the task, whereas those who had not learned the transition structure relied on the model-free learning strategy regardless of reward volatility. Reinforcement learning simulations revealed that the relative advantage of the model-based learning strategy over the model-free learning strategy peaked at intermediate levels of reward volatility. Additionally, increased time pressure shifted behavior toward the model-free learning strategy. These results demonstrated that, humans do not always use the model-based reinforcement learning strategy in uncertain and dynamic environments, even when they are aware of the task structure, supporting cost-benefit optimization. Author SummaryThe ability to flexibly guide behavior by carefully considering future consequences is fundamental to a prominent property of human intelligence and rationality. However, what drives this deliberative system? In this study, we investigated the factors that promote deliberative versus habitual behavior using decision-making tasks with uncertain structures and changing rewards. We found that participants who spontaneously learned the hidden transition structure in the task used this knowledge to guide deliberative behavior. Conversely, participants who did not learn the structure relied primarily on habitual strategies, repeating actions that had previously been rewarded. Among participants who learned the structure, the degree of deliberative behavior changed nonlinearly with reward volatility, in which the speed at which rewards changed over time. We also observed that limiting the decision time reduced deliberative behavior and promoted habitual responding. These findings suggest that under uncertain and dynamic environments, deliberative control is adaptively regulated according to cost-benefit optimization. Our results contribute to understanding how humans flexibly adjust their behavioral control systems in response to environmental conditions.

Where and how we move our eyes through a natural scene depends jointly on the scene and on the viewer. How the spatial and temporal organization of viewing changes across the lifespan, and whether those changes relate to memory, remains unclear. We recorded eye movements from a lifespan cohort (N = 179, ages 5-79) during free viewing of naturalistic scenes, then tested recognition across graded levels of image degradation. Characterizing each observer by how closely their viewing corresponded to that of age peers, young adults, and a stimulus-driven salience model, we found a developmental dissociation: consistency in where the eyes were directed increased monotonically with age, whereas consistency in how they moved -- saccade direction, length, and fixation duration -- followed an inverted-U peaking in young adulthood. Recognition sensitivity followed an inverted-U of the same form. Across all three reference frames, typicality in how the eyes moved, but not in their spatial targeting, predicted recognition.

Previous work showed that many participants fail to notice a gorilla in a video of people playing basketball. Another study found that 83% of trained radiologists failed to report a gorilla figure inserted into a chest CT nodule-search task, even though eye-tracking revealed that most observers had foveated the figure. We ask whether a similar phenomenon exists in contemporary vision-language models (VLMs). We find that (i) VLMs are capable of spotting the gorilla in both still-frame images and videos of lung CT scans; (ii) models display inattentional blindness, which varies according to model generation and type of stimulus presented; (iii) Gemini-3.1-Pro outperforms most other flagship and open-weight VLMs at identifying the presence or absence of the gorilla. We additionally ran a segmentation experiment utilizing two different model classes: a generalist (SAM 3), which found the gorilla but produced little to no results for anatomy-based prompts; a medical specialist (BiomedParse), which produced more promising anatomy-based results but flagged "gorilla" on gorilla-free control videos on 82% of frames. The behavioral signature of inattentional blindness reproduces in VLMs, but a unique confabulation failure mode means that any "did the model see X" claim requires signal-detection analysis with a matched-control false-alarm baseline.

The emotional content of suicide notes is typically examined using categorical coding, where each labeled passage is treated in isolation from its surrounding language. In contrast, dimensional models of psychopathology propose that affective content varies along continuous gradients. We evaluated this proposition directly. Excerpts from 884 annotated suicide notes were embedded in a semantic space defined solely by their linguistic properties, and we investigated whether human-assigned emotion labels changed smoothly across this space. They did: affective tone showed clear spatial autocorrelation (Moran's $I = 0.18$, $z = 19.68$, $p < 0.001$), an effect that replicated across three different encoders and remained after removing all within-note dependencies. Emotions occupied recognizable yet overlapping regions rather than forming distinct clusters and varied substantially in how tightly they were concentrated: love and hopelessness appeared with similar frequency, but love was far more localized ($z = 15.7$ versus $10.8$). Among all emotions, hopelessness was the most linguistically diffuse, implying that a single categorical label is capturing multiple, qualitatively different manifestations of suicidal distress.

Background General-purpose large language models increasingly encounter emotional and therapy-like conversation, yet are not developed or evaluated as clinical systems. Existing safety evaluations rely largely on brief exchanges, although harms often unfold over extended interactions. Whether models maintain safety-relevant performance as conversations accumulate context remains unknown. Methods In this preregistered study, 400 clinician-validated statements, with or without suicidal ideation, were inserted at 0-200 speaker turns in 5 psychotherapy and 3 synthetic transcripts. Forty-nine LLMs and 8 clinicians performed the same binary classification task. Mixed-effects models estimated the effects of conversational depth, model scale, and model version on F1. Twelve top models were tested to 1,500 turns across conversational trajectories, with or without instruction restatement. Results F1 declined with depth across model families (p<0.001). Larger, newer models performed better but still degraded. Clinicians showed no decline (mean F1 0.86 at both 0 and 200 turns), but eight of nine proprietary models exceeded their performance at 200 turns. Conversational content, not length alone, explained F1 changes; the largest decrease was under adversarial context (p<0.001). Restating instructions increased F1 on therapy to near baseline (median {Delta}F1 +0.12; p<0.001; 89% median recovery) versus MSJ ({Delta}F1 +0.08; p=0.04; 38% recovery). Conclusions LLM detection of suicidal ideation degraded with conversational depth and trajectory, whereas clinician performance remained stable despite the strongest models exceeding most clinicians in absolute performance. Mental health AI safety evaluations should test sustained performance across realistic and adversarial trajectories rather than relying on short-prompt benchmarks.

Suicidal thoughts and behaviours (STBs) are heterogeneous in their proximal dynamics, planning, and stress-sensitivity, yet most subtyping efforts remain symptom-driven and rarely validated across independent datasets. Computational mixture modelling offers a principled alternative: by fitting explicit models of learning and action selection and partitioning individuals by their latent parameter profiles, it can identify mechanistically distinct control strategies invisible to cross-sectional symptom measurement. We applied this approach to aversive Go/NoGo performance, jointly clustering two independently collected STB-enriched samples (N = 50 and N = 184) using tasks with the same structure but different duration, reversal timing, and clinical instrumentation. Two recurrent behavioural regimes emerged: a fast/adaptive regime characterised by rapid policy updating and elevated feedback reactivity, and a slow/perseverative regime characterised by slow updating, high choice determinism, and a pronounced cost following contingency reversal. These regimes were stable across initialisations, recovered more parsimoniously in joint than independent solutions, and were largely orthogonal to symptom-based stratification. Critically, stratification by regime exposed clinical-computational coupling structures substantially attenuated in pooled analyses. Pooled, population-level associations were modest and anchored by a broad affective burden axis. Within the slow/perseverative regime, coupling reorganised around learning dynamics and internalizing burden (depression, hopelessness, and active suicidal ideation) with markedly larger effect sizes. Within the fast/adaptive regime, a dissociation between anxious-compulsive and antisocial-disinhibitory profiles emerged along the same computational axis, invisible at the population level. These findings support a view of suicidality heterogeneity in which clinically similar individuals differ in the control strategies they recruit under aversive uncertainty - variation that symptom measurement alone cannot capture.