Psychological Review

We study how confidence in perceptual decisions depends on whether it is communicated verbally (e.g., "very likely") or numerically (e.g., "80% certainty"). We find that verbal expressions more reliably distinguish correct from incorrect choices than numerical reports, challenging the common assumption that numerical probabilities provide more precise representations of uncertainty. Additionally, in a dyadic decision-making task in which participants can revise their initial reports based on a partners choice and expressed confidence, verbal and numerical reports are equally effective in supporting accurate revisions of initial judgments. Together, these results underscore the effectiveness of verbal expressions as a means of conveying decision confidence.

The willingness to exert cognitive effort is essential but is constrained by the subjective cost of effort. Although effortful tasks are often avoided, positive bias about ones own performance may help sustain engagement with cognitive demands. Here, participants completed an effort-based decision-making task and reported trial-by-trial predictions of their own performance, allowing us to quantify performance prediction error (PPE) as the discrepancy between subjective and objective accuracy. The results showed that PPE was predominantly positive and increased with effort level, indicating greater overestimation under higher cognitive demands. Using a computational model, we show that choices were best explained by a learning model in which rewarded trials accompanied by positive PPE decreased subsequent sensitivity to effort. A confidence-based control model did not provide a better account of choices, suggesting that this effect was better captured by positive performance bias than by confidence alone. Our findings provide a computational account of how biased self-evaluation may attenuate the subjective cost of cognitive effort and extend the positive bias literature to the task need for cognitive effort.

Confirmation bias impacts judgments and decisions across a range of domains including finance, policy and science. Here we examine whether explicitly labelling information as true or false disrupts a core underlying computational mechanism that can generate this pervasive bias - asymmetric learning. Human participants (Study 1: N=47; Study 2: N=57) completed a 2 alternative forced choice (2AFC) task previously used to test for the presence of confirmation bias. Participants made choices between pairs of options that could win or lose money and received either factual or counterfactual feedback after each choice. We introduced a key novel feature into the task - providing explicit cues that signalled to participants whether feedback they had seen was true (verified) or false (debunked). Learning in response to feedback was attenuated under false compared to true labels but was present under both. Fitting participants choices to computational models enabled us to examine how sensitivity to the feedback varied as a function of both the label (true/false) and confirmation (confirmatory/disconfirmatory). This revealed a distinct pattern of learning rates typical of confirmation bias (enhanced learning from positive prediction errors for chosen options and from negative prediction errors for unchosen options) in response to both true and false labels. The findings highlight how confirmation bias plays an important role in the effectiveness of interventions designed to verify true and/or debunk false claims. Verification is less likely to succeed when information disconfirms prior beliefs. Conversely, debunking false claims is unlikely to succeed when the information confirms ones prior beliefs.

Saccades made during memory maintenance prioritize memory for the saccade target, but it is unclear if this benefit is specific to a location or extends across memorized objects. In three experiments, we examined whether saccadic selection spreads to other locations within the same object. In Experiment 1, we asked observers to remember three oriented Gabors presented either within contour-defined objects or without object structure. A subsequent movement cue prompted observers to move their eyes to the indicated location. We then probed memory for stimuli at locations equidistant from the saccade target, in either the same or a different object. Memory was best for stimuli at locations congruent with the saccade target, and consistently weaker for other stimuli presented in the same or a different object than the saccade target. In Experiment 2, we created more complex objects by adding more object features to the stimulus. Again, memory performance was best for stimuli congruent with the saccade target location, whereas memory in incongruent trials was worse and similar for stimuli in the same and different object as the saccade target. In Experiment 3, we tested if saccadic selection is present and propagates within the object in a change detection task. Again, memory performance (i.e., change detection) was best at the saccade target location. However, this memory benefit also spread to other locations within the same object. Our results imply that saccadic selection in visual working memory is primarily space-based but can also spread towards locations within the object where a saccade was directed.

Goal-directed behavior often requires sustained effort across a sequence of interdependent decisions, yet the determinants of persistence in such contexts remain poorly understood. Here, we investigated how individuals regulate persistence in a novel sequential effort-based task in which they controlled an avatar through successive checkpoints to reach a final goal and could make repeated attempts following failure. At each attempt, participants could choose either to persist in the same task or to disengage toward an easier but less rewarding alternative. We found that decisions to persist or disengage were jointly shaped by multiple interacting factors. Disengagement increased with task difficulty and lower skill level. It also increased with repeated attempts and time-on-task, indexing fatigue, and with accumulated errors, indexing lack of progress. Conversely, proximity to the goal promoted persistence and shaped decision dynamics by reducing choice conflict during persistence decisions and increasing hesitation during disengagement near the goal. Notably, clearing the first checkpoint produced a sharp increase in persistence, suggesting that early success plays a pivotal role. Furthermore, persistence reflected both retrospective and prospective evaluations of effort, with prior investment promoting commitment and anticipated effort reducing it. Finally, disengagement was preceded by short-term performance decline but not by gradual increases in decision conflict, suggesting relatively abrupt strategy shifts following repeated failures. Together, these findings provide a comprehensive account of persistence in sequential effortful tasks, showing that decisions to persist or disengage are jointly shaped by multiple factors related to fatigue, (lack of) progress, goal proximity, and early success.

When humans learn under conditions of uncertainty, they dynamically adjust how they prepare for and respond to feedback. In navigating uncertain environments, the brain minimizes error by continuously refining internal models via memory updating (MU). Feedback is critical for MU, and anticipatory neural mechanisms shape how feedback is processed, likely reflecting learned environmental certainty. However, the literature has largely focused on post-feedback activity, leaving pre-feedback certainty-related mechanisms less understood. The present study aims to address this gap by examining how certainty modulates anticipatory states, preceding feedback and subsequent MU. We examined oscillatory activity prior to performance feedback in a reanalysis of EEG data previously published by Hassall and colleagues (2023). Twenty-one participants (16 female, Mage = 25.81 years) predicted the strength of cartoon characters with varying predictability levels which were learned through exposure. Feedback on prediction accuracy was presented via an animated rising bar. Results revealed that theta power is modulated by accumulative feedback. Linear mixed-effects models revealed an interaction between predictability-related certainty and learning stage: in late learning, higher performance was associated with increased pre-feedback alpha and beta power for low-certainty trials, whereas in early learning, higher performance was associated with decreased beta power. These learning-related modulations in alpha and beta power suggest that initial learning is marked by adaptable exploratory processing. Subsequent learning exhibited increased alpha-mediated inhibition and beta-related anticipatory activity for lower certainty trials, indicative of dynamic strategy refinement and selective engagement of task-relevant information. These results demonstrate that certainty shapes preparatory oscillatory activity associated with MU.

Visual working memory (vWM) is often linked to conscious experience and visual imagery, but it is typically described as a system that stores separate, independent items. These assumptions are difficult to reconcile, given the unified nature of conscious experience. Here, we test the hypothesis that vWM relies on at least two distinct representations: an underlying, unconscious memory trace and a consciously accessible, integrated representation. A total of 216 participants performed a change-detection task, in which they rated their perceptual awareness of the memory display during the maintenance interval. Critically, we manipulated the statistical properties of the displays (average item size and size variability) to probe sensitivity to unified ensemble-level structure. Results revealed a dissociation between subjective and objective measures. Perceptual awareness increased for displays with larger, more variable items, whereas objective performance improved for displays with smaller, less variable items. Despite this difference, subjective awareness still predicted performance, and even incorrect responses showed consistent biases rather than random guesses. Importantly, individual differences in imagery vividness (VVIQ) were selectively associated with subjective awareness and estimation bias, but not with objective correctness. These precision biases were further shaped by display statistics, suggesting that multiple representations can guide behavior. Together, our findings support a reinterpretation of vWM performance in which task responses can draw on both unconscious and consciously accessible representations. One possible explanation for these behavioral patterns is that subjective experience reflects integrated, ensemble-like representations, while objective performance depends more strongly on item-specific information. Public significance statementsWorking memory allows us to temporarily hold and use information, and differences in this ability are closely linked to broader cognitive skills such as intelligence. This study shows that these differences may not depend only on how much information people can store, but also on how they experience it: some individuals appear to rely more on consciously accessible, image-like representations, especially when memory is uncertain or prone to error. By demonstrating that subjective experience and the vividness of imagery can shape behavior independently of objective accuracy, these findings suggest that how we use memory may be as important as how much we can store, with implications for understanding individual differences in cognition.

Real-world distractors occur in environments whose states change at different rates. We asked whether such volatility alters early attentional gating or instead changes the criterion for committing to a response. Observers performed an additional-singleton search task with concurrent eye tracking while distractor presence followed high- or low-volatility sequences, with overall distractor prevalence held constant. Trial-pooled oculomotor capture was higher under high volatility, a pattern that appears to indicate altered filtering. That inference did not survive repetition-aware analysis: once the same-location run position was matched, capture did not detectably differ across volatility regimes. The pooled capture effect was therefore consistent with a structural consequence of the volatility manipulation, which enriched high-volatility blocks with early-run positions where capture is intrinsically high. The positive volatility signature appeared on distractor-absent trials, where high-volatility blocks were associated with longer target latency, more fixations, longer final-target dwell, and fewer errors. Same-location repetition learning showed no detectable difference in slope across regimes. A hierarchical drift-diffusion model (DDM) and a complementary volatility Kalman-filter (VKF) dynamic-state comparison indicated that manual responses were better described by architectures that allow both boundary-related and drift-related components than by a boundary-only account. Volatility, therefore, did not show detectable evidence of impairing the local gating rule; instead, the converging evidence points to a post-selective verification/caution profile, consistent with a precision-weighted read-out of environmental uncertainty.

Internal forward models predict the sensory consequences of motor commands; however, whether the anticipated availability of post-action feedback contributes to the precision of the action itself remains unknown. We manipulated the predictability of post-release visual occlusion in skilled basketball players. Participants performed three-point shots while wearing liquid-crystal shutter goggles. The study tested three conditions: a no-occlusion baseline, certain-occlusion condition in which players knew that their vision would be occluded at ball release in every trial, and random-occlusion condition in which they could not predict whether an occlusion would occur. Shooting accuracy declined in the certain-occlusion condition relative to the no-occlusion condition (49.2% vs 41.7%). The random-occlusion condition did not differ from the baseline (46.1%). Within the random condition, the accuracy in occluded trials were virtually identical to that in non-occluded trials (46.6% vs 46.2%), even though the immediate visual occlusion was the same as in the certain-occlusion condition. These results demonstrate that it is not the absence of post-action information per se that disrupts motor execution, but the prior certainty that action consequences will be unavailable. We interpret this finding as a prospective influence of anticipated consequence loss, whereby motor execution depends on whether the prediction-outcome loop remains closable.

In contrast to static formalisms, computational definitions describe the operational mechanisms of a model. Simulations are an essential part of the cycle of theory development and refinement, assisting researchers in formulating the precise definitions that models require, and making accurate predictions. This manuscript introduces a computational implementation of Pavlovian learning models in a Python environment, termed Pavlovian Associative Learning Models Simulation (PALMS). In addition to the canonical Rescorla-Wagner model, attentional approaches are implemented, including Pearce-Kaye-Hall, Mackintosh Extended, Le Pelleys Hybrid, and a novel extension of the Rescorla-Wagner model featuring a unified variable learning rate that synthesises Mackintoshs and Pearce and Halls opposing conceptualisations. To our knowledge, only the first attentional model has been previously specified computationally in a general design tool. PALMS integrates a graphical interface that permits the input of entire experimental designs in an alphanumeric format, akin to that used by experimental neuroscientists. It uniquely enables the simulation of experiments involving hundreds of stimuli, such as those used with human participants, and the computation of configural cues and configural-cue compounds across all models, thereby substantially broadening their predictive capabilities. A comprehensive description of the models implementation and the environment functionalities is provided in the paper; these include efficient and accurate operation and instant visualisation of predicted results across different models within a single architecture and environment. We evaluate PALMS by simulating five published experiments in the associative learning literature that assessed the predictive scope of existing models, and we show that this implementation provides neuroscientists with a useful tool for identifying critical variables, refining experimental designs, making precise predictions, comparing model fitness, and formulating new theoretical approaches. PALMS is licensed under the open-source GNU Lesser General Public License 3.0. The environment source code and the latest multiplatform release build are accessible as a GitHub repository at https://github.com/cal-r/PALMS-Simulator. Author summaryResearch on associative learning is multidisciplinary, encompassing disciplines such as neuroscience, AI, psychology, psychiatry, behavioural sciences, planning, and marketing. Unlike static formalisms, precise computational definitions specify how a model operates, enabling model simulation, swift and error-free prediction calculations, which are essential for testing theories, comparing predictions, holding models accountable, and providing a common language across fields. We introduce Pavlovian Associative Learning Models Simulation (PALMS), a user-friendly, open-source Python environment for simulating classical conditioning and studying the role of attention in learning. PALMS implements the prescriptive Rescorla-Wagner and attentional models: Pearce-Kaye-Hall, Mackintosh Extended, Le Pelleys Hybrid, and a new hybrid model with a unified variable learning rate that blends Mackintosh and Pearce-Halls conflicting views. Its graphical interface makes it easy for neuroscientists to enter experiments. Our computational implementation supports simulations with hundreds of stimuli, configural cues, and compounds, broadening the models predictive power. Designed for efficiency, it offers instant visual results and useful features. We evaluate PALMS by simulating five published experiments, highlighting its value for model comparison and refinement, and, more generally, as a tool to assist research.

RationaleCholinergic signalling is critical for attentional control and signal detection, yet the contribution of specific acetylcholine receptor (AChR) subtypes remains poorly understood. Although the 7 nicotinic AChR (nAChR) holds promise as a target for cognition-enhancing therapy, clinical findings to date have been inconsistent. ObjectiveTo investigate the effects of putative cognitive enhancing drugs, including those targeting cholinergic transmission and 7 nAChRs on a visual signal detection task (SDT). MethodsMale and female Sprague Dawley rats were trained on an SDT. Cholinergic transmission was probed systemically with nicotinic and muscarinic receptor antagonists (mecamylamine and scopolamine), a cholinesterase inhibitor (galantamine), an M4-AChR positive allosteric modulator (PAM; VU0467154), an 7 nAChR antagonist (MLA), an 7 nAChR PAM (CCMI), and an 7 nAChR partial agonist (SSR-180,711). Dopaminergic transmission was probed using the catechol-O-methyltransferase (COMT) inhibitor, tolcapone. A novel, trial-level signal detection theory-based generalised linear mixed-effects model (SDT-GLMM) was used to index response bias and perceptual sensitivity (d'), the latter reflecting subjects ability to discriminate signal from noise. ResultsMecamylamine profoundly impaired SDT performance across all measures. Galantamine significantly improved d' at moderate doses but not when a distractor was present. MLA uniquely produced dose-dependent improvements in d' that were preserved under distraction. In contrast, positive allosteric modulation and agonism of 7 nAChRs impaired task performance. Scopolamine, VU0467154, and tolcapone had no consistent or interpretable effects on signal detection. ConclusionsThis work demonstrates that 7 nAChR modulation bidirectionally and dose-dependently regulates perceptual sensitivity, irrespective of attentional distraction. These findings have implications for targeted cognitive enhancement in disorders of attention.

Subthreshold depression is associated with significant functional impairment and elevated risk of major depressive disorder. A negative self-concept may disrupt the implicit positive association evoked by ones own face, impairing incidental encoding of self-relevant information. Whether subthreshold depression involves a selective deficit in encoding self-face identity remains unclear. The attribute amnesia paradigm is well suited to address this question because it can dissociate attentional selection from working memory encoding. Using this paradigm, we examined the issue across two experiments. Experiment 1 employed nonsocial stimuli (animal drawings) and confirmed an intact attribute amnesia effect in subthreshold depression (n = 30) comparable to healthy controls (n = 30), ruling out a generalized encoding deficit. Experiment 2 replaced targets with faces (self or other) and revealed a selective enhancement of the attribute amnesia effect for self-face identity in subthreshold depression. Specifically, on the surprise trial, accuracy for self-face identity dropped to near-chance levels in the subthreshold depression group, whereas no such deficit emerged for other-faces or in controls. Encoding recovered rapidly once explicit memory expectations were introduced, indicating intact basic encoding capacity. These findings suggest that subthreshold depression is associated with a specific impairment in incidentally encoding self-face identity. This impairment likely stems from a negative self-concept that weakens self-face salience under incidental encoding conditions. By capturing this selective encoding failure, the present study reveals that the self-processing deficit in subthreshold depression can arise at the gating stage between attention and working memory consolidation.

Humans comprehend language incrementally, updating the representation of sentence meaning with each incoming word. These updates are guided by the distance between each perceived word and prior expectations--the prediction error. The alignment between large language models (LLMs) and cortical activity inspires the hypothesis that the cortical computation of prediction error is Surface-based, driven by statistical patterns of word form co-occurrence. In contrast, psycholinguistic models propose that prediction error computation is Meaning-based, driven by word semantics. We used polysemic words with ambiguous semantics to distinguish these models: ambiguity would introduce uncertainty into meaning representations and hence the prediction error, if Meaning-based, but would not affect the prediction error, if Surface-based. We examined how ambiguity influenced prediction error signatures in self-paced reading times and magnetoencephalographic (MEG) neural responses during sentence processing. While an LLM-based proxy of prediction error robustly predicted reading times and neural responses to unambiguous words, it failed to predict either under ambiguity. That is, prediction error computation was altered by uncertainty in word meaning, which supports the Meaning-based model and corroborates the essential role of word meaning in predictive language processing. Our findings highlight an important limitation of LLMs as in silico models of the human language faculty.

Population analysis methods have become standard for navigating the complexity of neural data. However, these methods often assume a rate code, neglecting information encoded in the precise timing of spikes. Critically, additional information encoded in bursts of action potentials may be missed. Here, we develop a factor analysis method that disentangles the factors associated with bursts and individual spikes. This enables burst codes to be investigated directly from the structure of the data, without requiring external covariates. We demonstrate that analyzing firing rates alone obscures the latent structure and factors underlying bursts. Applying our method to simulated and experimental data, we show that it can infer the correct latent structure and be used to test for the presence of burst coding. By merging the population and burst coding perspectives, we provide a framework for linking changes in bursting to internal variables involved in attention, perception, and learning.

Humans frequently make decisions in complex, high-dimensional environments, where identifying task-relevant information is critical for rapid behavior optimization. Humans outperform standard reinforcement learning agents in navigating such complexity, yet the cognitive strategies of humans remain unclear. To address this, we developed a novel multi-dimensional learning task in which only a subset of dimensions is reward-related. Crucially, unlike prior studies, subjects are uninformed of the true task dimensionality and have to identify them through exploration. This design closely mimics the ambiguity in real-world tasks. Our results have identified two stereotyped choice patterns that reveal "dimension-guided" strategies in exploration and exploitation. Cross-subject analyses suggest that dimension-guided exploration may promote the efficiency of reward-based learning. These findings indicate that humans leverage task dimensionality to guide exploration, and provide inspiration for improving exploration efficiency in AI agents.

Human goal pursuit is often marked by persistent activity toward achieving an objective, as well as flexibility in switching objectives based on environmental demands. How humans balance the stability and flexibility necessary for goal pursuit is the key question of this study. We propose that goal pursuit generates dynamic attractor modes in policy landscapes that produce stability in goal pursuit. The attractor properties are modulated through progress monitoring, allowing for the flexibility necessary to switch objectives in favor of alternative goals. Through simulations and behavioral cloning of human participants performing an extended goal selection task, we show how dynamic modes can develop in the latent spaces of recurrent neural networks trained with reinforcement learning. We develop metrics to quantitatively assess the attractor qualities of dynamic modes, validating them against synthetically generated dynamical systems, and use them to investigate the context modulation of attractor modes during goal pursuit. We then proceed to develop a circuit-level account of goal persistence incorporating self-excitation and cross-inhibition as motifs for fast, self-sustaining dynamics modulated by slow, progress-integrating momentum and context signals. Lastly, we show that the switching costs experienced while managing multiple goals are an emergent property of resistance to the intrinsic dynamics of goal pursuit, thereby contributing a fresh perspective on the dynamics of extended goal pursuit.

IntroductionAlthough neural activity is organized across temporal and spatial scales, the principles that determine the accuracy and fidelity of neural information representation across scales remain unclear. In particular, while recent empirical results have reported mesoscopic optimality in neural decoding, no theoretical accounts exist that explain when and why such intermediate scales emerge as optimal. Here, we develop an analytical framework to study the optimal temporal scale of information representation and its dependence on the dynamic structure of signal and noise in neural data. Materials and MethodsWe formulate a multiscale theoretical model in which neural population activity is represented by temporally encoded trial vectors at microscopic, mesoscopic, and macroscopic resolutions. Neural responses are modeled as class-dependent mean activations (signal) corrupted by temporally correlated noise, and decay rates of correlations in both signal and noise are varied parametrically. Representational quality at each scale is quantified using the sensitivity index (d-prime) for decoding condition from neural activity. ResultsWe derive closed-form expressions for the sensitivity index at each temporal scale. These expressions reveal the key roles of signal and noise correlations as the main determinants of condition decodability at all scales. Comparing expressions under various combinations of signal and noise correlations reveals two regimes. First, when signal and noise correlations are absent or persistent over time, the optimal resolution falls at one of two extremes: macroscale (resp. microscale) if signal correlations are stronger (resp. weaker) than noise correlations. In contrast, when both signal and noise correlations decay with temporal separation, temporal integration produces a nontrivial trade-off: moderate integration improves decodability by suppressing noise while preserving coherent signal, whereas excessive integration degrades signal and decodability. Therefore, only in the latter regime, mesoscopic representations emerge as optimal across a broad range of biologically plausible parameters. DiscussionThis work provides a theoretical explanation for how the optimal temporal scale of neural information representation depends on the interplay between signal and noise correlations and their temporal decay. Broadly, the framework establishes temporal integration as a principled mechanism linking multiscale neural dynamics to information representation and offers testable predictions across recording modalities and neural systems.

Perceptual systems in humans and many other animals are able to segment scenes into regions that are likely to be physically meaningful. This ability depends on having low-level mechanisms that can accurately categorize whether local image patches are samples from the same or different kinds of texture. We find that using spatial proximity as a proxy for same-different ground truth makes it possible to train accurate decision variables and bounds directly from arbitrary natural images with no feedback. We also find that performance can be further improved by using proximity as a ground truth for adjusting the final decision variables and bounds for the current image/scene. These surprising findings result from the simple fact that under a wide range of conditions proximity discrimination (near vs. far) and texture discrimination (same vs. different) have mathematically identical decision bounds if the same image features are used for both tasks. We used the decision variables and bounds trained on natural images as the initial steps in a hierarchical Bayesian observer (HBO) model of texture discrimination [9]. Given the relative simplicity of this HBO model, it did an excellent job of segmenting images having randomly shaped regions containing arbitrary natural textures. We suggest that the proximity proxy is something that natural selection could discover and exploit for any same-different task where the task-relevant stimulus features also vary systematically with distance in space and/or time. For example, natural selection could have created developmental learning/plasticity mechanisms that exploit the proximity proxy.

Summary paragraphA hallmark of learning is the need for sensory stimuli (Ginns, 2015; McGraw et al., 2009; Reinwein, 2012; Spence, 1950) so that learning is fundamentally based on sensory input signals affecting behaviour, physiology, and neurology. If behavioural measures of learning can be causally linked to physiological and neurological variables, a broader understanding of the mechanisms related to learning in schools, learning disabilities, and learning and health issues may emerge (McGraw et al., 2009). Despite decades of research on the physiological/neurological variable of sympathetic activation, learning, and achievement (Horvers et al., 2021), any causal relation remains unclear (Cowley et al., 2014; Mason et al., 2020; Pijeira-Diaz et al., 2016; Sung et al., 2023; Yu et al., 2024) and issues with instrument validation remain (Costantini et al., 2023; Hu et al., 2024; Milstein & Gordon, 2020; Van Der Mee et al., 2021). Here we investigate the effect of sensory input on sympathetic activation by using validated instruments for skin conductance measurement (Batista et al., 2019) and whether sympathetic activation is connected to learning in a cognitive laboratory context and an ecologically valid classroom context. In both contexts, we found a physiological variable which correlated with learning and that sensory input affected this variable while student movement did not. These sensory inputs varied depending on the different instructional activities the students participated in. Together, these findings bring us one step closer to a model linking sensory input to behavioural, physiological, and neurological variables.

How do humans store sequences that far exceed working memory capacity? Using visuo-spatial and binary auditory sequences, we previously showed that a Language of Thought (LoT) architecture -- in which simple primitives are recursively combined into hierarchical programs -- enables efficient storage of structured sequences. Here we ask whether this principle extends to purely ordinal structure: sequences defined by how items repeat and in what order, as in AABBCCAABBCC, independently of their spatial content. Across three experiments, participants reproduced 12-item sequences of spatial locations with various ordinal structures. The minimal description length derived from the LoT model predicted recall accuracy with remarkable precision (r = .96), substantially outperforming Shannon entropy, Lempel-Ziv complexity, chunking models and subjective complexity ratings. Critically, fine-grained analyses of participants inter-click intervals during reproduction revealed systematic slowdowns at the hierarchical boundaries predicted by the LoT programs, providing a behavioral signature of the underlying mental syntax. These results identify a compact vocabulary of mental primitives -- repetition, mirroring, and interleaving -- whose composition accounts for the symbolic compression of ordinal structures. For ordinal regularities, human sequence memory operates as a form of program induction, leveraging a domain-general capacity for hierarchical compression to encode complex structured information. Author SummaryHuman short-term memory is heavily limited, holding no more than a few items at once. Yet humans routinely memorize complex sequences that far exceed this capacity. How is this possible? We propose that the brain acts like a programmer: rather than storing each element individually, it compresses sequences into short mental "programs." Just as a programmer writes "repeat ABC four times" instead of typing ABCABCABCABC, the brain leverages regularities such as repetitions (ABC-ABC) or mirror patterns (ABC-CBA) to encode sequences efficiently. We tested this idea across three experiments: two in which participants memorized and reproduced sequences of spatial positions on a screen, one where they only rated their perceived complexity. Sequences described by shorter programs were remembered far better and judged as simpler -- even when they were the same length as less structured sequences. When reproducing sequences, participants paused longer at structural boundaries, revealing the internal organization of their mental programs. Strikingly, program length predicted memory performance better than participants own complexity ratings, suggesting that these mental representations are not fully accessible to conscious awareness. Finally, we identified key new patterns -- including temporal inversion and interleaving -- that extend the Language of Thought framework. Together, these findings suggest that a compositional Language of Thought is a fundamental aspect of how the human brain efficiently store and represent structured information.