Infants tailor their attention to maximize learning

Forward and backward blocking in statistical learning

Prediction errors have a prominent role in many forms of learning. For example, in reinforcement learning, agents learn by updating the association between states and outcomes as a function of the prediction error elicited by the event. One paradigm often used to study error-driven learning is blocking. In forward blocking, participants are first presented with stimulus A, followed by outcome X (A→X). In the second phase, A and B are presented together, followed by X (AB→X). Here, A→X blocks the formation of B→X, given that X is already fully predicted by A. In backward blocking, the order of phases is reversed. Here, the association between B and X that is formed during the first learning phase of AB→X is weakened when participants learn exclusively A→X in the second phase. The present study asked the question whether forward and backward blocking occur during visual statistical learning, i.e., the incidental learning of the statistical structure of the environment. In a series of studies, using both forward and backward blocking, we observed statistical learning of temporal associations among pairs of images. While we found no forward blocking, we observed backward blocking, thereby suggesting a retrospective revaluation process in statistical learning and supporting a functional similarity between statistical learning and reinforcement learning.

Toddlers strategically adapt their information search

Adaptive information seeking is essential for humans to effectively navigate complex and dynamic environments. Here, we developed a gaze-contingent eye-tracking paradigm to examine the early emergence of adaptive information-seeking. Toddlers (N = 60, 18-36 months) and adults (N = 42) either learnt that an animal was equally likely to be found in any of four available locations, or that it was most likely to be found in one particular location. Afterwards, they were given control of a torchlight, which they could move with their eyes to explore the otherwise pitch-black task environment. Eye-movement data and Markov models show that, from 24 months of age, toddlers become more exploratory than adults, and start adapting their exploratory strategies to the information structure of the task. These results show that toddlers' search strategies are more sophisticated than previously thought, and identify the unique features that distinguish their information search from adults'.

Breaking down (and moving beyond) novelty as a trigger of curiosity

The Novelty Seeking Model (NSM) places "novelty" at center stage in characterizing the mechanisms behind curiosity. We argue that the NSM's conception of novelty is too broad, obscuring distinct constructs. More critically, the NSM underemphasizes triggers of curiosity that better unify these constructs and that have received stronger empirical support: those that signal the potential for useful learning.

A shared novelty-seeking basis for creativity and curiosity: Response to the commentators

In our target article, we proposed that curiosity and creativity are both manifestations of the same novelty-seeking process. We received 29 commentaries from diverse disciplines that add insights to our initial proposal. These commentaries ultimately expanded and supplemented our model. Here we draw attention to five central practical and theoretical issues that were raised by the commentators: (1) The complex construct of novelty and associated concepts; (2) the underlying subsystems and possible mechanisms; (3) the different pathways and subtypes of curiosity and creativity; (4) creativity and curiosity "in the wild"; (5) the possible link(s) between creativity and curiosity.

Curiosity and the dynamics of optimal exploration

What drives our curiosity remains an elusive and hotly debated issue, with multiple hypotheses proposed but a cohesive account yet to be established. This review discusses traditional and emergent theories that frame curiosity as a desire to know and a drive to learn, respectively. We adopt a model-based approach that maps the temporal dynamics of various factors underlying curiosity-based exploration, such as uncertainty, information gain, and learning progress. In so doing, we identify the limitations of past theories and posit an integrated account that harnesses their strengths in describing curiosity as a tool for optimal environmental exploration. In our unified account, curiosity serves as a 'common currency' for exploration, which must be balanced with other drives such as safety and hunger to achieve efficient action.

Individual differences in processing speed and curiosity explain infant habituation and dishabituation performance

Habituation and dishabituation are the most prevalent measures of infant cognitive functioning, and they have reliably been shown to predict later cognitive outcomes. Yet, the exact mechanisms underlying infant habituation and dishabituation are still unclear. To investigate them, we tested 106 8-month-old infants on a classic habituation task and a novel visual learning task. We used a hierarchical Bayesian model to identify individual differences in sustained attention, learning performance, processing speed and curiosity from the visual learning task. These factors were then related to habituation and dishabituation. We found that habituation time was related to individual differences in processing speed, while dishabituation was related to curiosity, but only for infants who did not habituate. These results offer novel insights in the mechanisms underlying habituation and serve as proof of concept for hierarchical models as an effective tool to measure individual differences in infant cognitive functioning. RESEARCH HIGHLIGHTS: We used a hierarchical Bayesian model to measure individual differences in infants' processing speed, learning performance, sustained attention, and curiosity. Faster processing speed was related to shorter habituation time. High curiosity was related to stronger dishabituation responses, but only for infants who did not habituate.

Curiosity: primate neural circuits for novelty and information seeking

For many years, neuroscientists have investigated the behavioural, computational and neurobiological mechanisms that support value-based decisions, revealing how humans and animals make choices to obtain rewards. However, many decisions are influenced by factors other than the value of physical rewards or second-order reinforcers (such as money). For instance, animals (including humans) frequently explore novel objects that have no intrinsic value solely because they are novel and they exhibit the desire to gain information to reduce their uncertainties about the future, even if this information cannot lead to reward or assist them in accomplishing upcoming tasks. In this Review, I discuss how circuits in the primate brain responsible for detecting, predicting and assessing novelty and uncertainty regulate behaviour and give rise to these behavioural components of curiosity. I also briefly discuss how curiosity-related behaviours arise during postnatal development and point out some important reasons for the persistence of curiosity across generations.

Playing hide and seek: Contextual regularity learning develops between 3 and 5 years of age

The ability to acquire contextual regularities is fundamental in everyday life because it helps us to navigate the environment, directing our attention where relevant events are more likely to occur. Sensitivity to spatial regularities has been largely reported from infancy. Nevertheless, it is currently unclear when children can use this rapidly acquired contextual knowledge to guide their behavior. Evidence of this ability is indeed mixed in school-aged children and, to date, it has never been explored in younger children and toddlers. The current study investigated the development of contextual regularity learning in children aged 3 to 5 years. To this aim, we designed a new contextual learning paradigm in which young children were presented with recurring configurations of bushes and were asked to guess behind which bush a cartoon monkey was hiding. In a series of two experiments, we manipulated the relevance of color and visuospatial cues for the underlying task goal and tested how this affected young children's behavior. Our results bridge the gap between the infant and adult literatures, showing that sensitivity to spatial configurations persists from infancy to childhood, but it is only around the fifth year of life that children naturally start to integrate multiple cues to guide their behavior.

Multifaceted information-seeking motives in children

From an early age, children need to gather information to learn about their environment. Deciding which knowledge to pursue can be difficult because information can serve several, sometimes competing, purposes. Here, we examine the developmental trajectories of such diverse information-seeking motives. Over five experiments involving 521 children (aged 4-12), we find that school-age children integrate three key factors into their information-seeking choices: whether information reduces uncertainty, is useful in directing action, and is likely to be positive. Choices that likely reveal positive information and are useful for action emerge as early as age 4, followed by choices that reduce uncertainty (at ~age 5). Our results suggest that motives related to usefulness and uncertainty reduction become stronger with age, while the tendency to seek positive news does not show a statistically significant change throughout development. This study reveals how the relative importance of diverging, sometimes conflicting, information-seeking motives emerges throughout development.

The Role of Talking Faces in Infant Language Learning: Mind the Gap between Screen-Based Settings and Real-Life Communicative Interactions

Over the last few decades, developmental (psycho) linguists have demonstrated that perceiving talking faces audio-visually is important for early language acquisition. Using mostly well-controlled and screen-based laboratory approaches, this line of research has shown that paying attention to talking faces is likely to be one of the powerful strategies infants use to learn their native(s) language(s). In this review, we combine evidence from these screen-based studies with another line of research that has studied how infants learn novel words and deploy their visual attention during naturalistic play. In our view, this is an important step toward developing an integrated account of how infants effectively extract audiovisual information from talkers' faces during early language learning. We identify three factors that have been understudied so far, despite the fact that they are likely to have an important impact on how infants deploy their attention (or not) toward talking faces during social interactions: social contingency, speaker characteristics, and task- dependencies. Last, we propose ideas to address these issues in future research, with the aim of reducing the existing knowledge gap between current experimental studies and the many ways infants can and do effectively rely upon the audiovisual information extracted from talking faces in their real-life language environment.

Eight-Month-Old Infants Meta-Learn by Downweighting Irrelevant Evidence

Infants learn to navigate the complexity of the physical and social world at an outstanding pace, but how they accomplish this learning is still largely unknown. Recent advances in human and artificial intelligence research propose that a key feature to achieving quick and efficient learning is meta-learning, the ability to make use of prior experiences to learn how to learn better in the future. Here we show that 8-month-old infants successfully engage in meta-learning within very short timespans after being exposed to a new learning environment. We developed a Bayesian model that captures how infants attribute informativity to incoming events, and how this process is optimized by the meta-parameters of their hierarchical models over the task structure. We fitted the model with infants' gaze behavior during a learning task. Our results reveal how infants actively use past experiences to generate new inductive biases that allow future learning to proceed faster.

Encoding interactive scripts at 10 months of age

Understanding action-reaction associations that give origin to interactive scripts (e.g., give-and-take interactions) is essential for appreciating social exchanges. However, studies on infants' action understanding have mainly investigated the case of actions performed by individual agents. Moreover, although extensive literature has explored infants' comprehension of action-effect relationships in object functioning, no study has addressed whether it also plays a role when observing social interactions, an issue we addressed here. In a first study, 10-month-old infants observed short videos of dyadic exchanges. We investigated whether they were able to link specific human gestures directed toward another person to specific vocal reactions in the receiver. We used a double-habituation paradigm in which infants were sequentially habituated to two specific action-reaction associations. In the test phase, infants watched one of the two habituated (Familiar) videos, a video with a reversed action-reaction association (Violation), and a Novel video. Results showed that the infants looked longer at both the Novel and Violation test trials than at the Familiar test trials. In a control study, we show that these results could not be accounted for by associative learning; indeed, learning of the action-reaction association did not occur when the vocalization was not produced by the receiver but only contingent on the agent's action. Thus, we show that 10-month-old infants can encode specific social action-effect relationships during the observation of dyadic interactions and that the interactivity of the social context may be critical to shaping young infants' understanding of others' behaviors.

Are infants' preferences in the number change detection paradigm driven by sequence patterns?

Inter-individual differences in infants' numerosity processing have been assessed using a change detection paradigm, where participants were presented with two concurrent streams of images, one alternating between two numerosities and the other showing one constant numerosity. While most infants look longer at the changing stream in this paradigm, the reasons underlying these preferences have remained unclear. We suggest that, besides being attracted by numerosity changes, infants perhaps also respond to the alternating pattern of the changing stream. We conducted two experiments (N = 32) with 6-month-old infants to assess this hypothesis. In the first experiment, infants responded to changes in numerosity even when the changing stream showed numerosities in an unpredictable random order. In the second experiment, infants did not display any preference when an alternating stream was pitted against a random stream. These findings do not provide evidence that the alternating pattern of the changing stream contributes to drive infants' preferences. Instead, around the age of 6 months, infants' responses in the numerosity change detection paradigm appear to be mainly driven by changes in numerosity, with different levels of preference reflecting inter-individual difference in the acuity of numerosity perception.

Curiosity as a metacognitive feeling

Curious information-seeking is known to be a key driver for learning, but characterizing this important psychological phenomenon remains a challenge. In this article, we argue that solving this challenge requires qualifying the relationships between metacognition and curiosity. The idea that curiosity is a metacognitive competence has been resisted: researchers have assumed both that young children and non-human animals can be genuinely curious, and that metacognition requires conceptual and culturally situated resources that are unavailable to young children and non-human animals. Here, we argue that this resistance is unwarranted given accumulating evidence that metacognition can be deployed procedurally, and we defend the view that curiosity is a metacognitive feeling. Our metacognitive view singles out two monitoring steps as a triggering condition for curiosity: evaluating one's own informational needs, and predicting the likelihood that explorations of the proximate environment afford significant information gains. We review empirical evidence and computational models of curiosity, and show that they fit well with this metacognitive account, while on the contrary, they remain difficult to explain by a competing account according to which curiosity is a basic attitude of questioning. Finally, we propose a new way to construe the relationships between curiosity and the human-specific communicative practice of questioning, discuss the issue of how children may learn to express their curiosity through interactions with others, and conclude by briefly exploring the implications of our proposal for educational practices.

How infant-directed actions enhance infants' attention, learning, and exploration: Evidence from EEG and computational modeling

When teaching infants new actions, parents tend to modify their movements. Infants prefer these infant-directed actions (IDAs) over adult-directed actions and learn well from them. Yet, it remains unclear how parents' action modulations capture infants' attention. Typically, making movements larger than usual is thought to draw attention. Recent findings, however, suggest that parents might exploit movement variability to highlight actions. We hypothesized that variability in movement amplitude rather than higher amplitude is capturing infants' attention during IDAs. Using EEG, we measured 15-month-olds' brain activity while they were observing action demonstrations with normal, high, or variable amplitude movements. Infants' theta power (4-5 Hz) in fronto-central channels was compared between conditions. Frontal theta was significantly higher, indicating stronger attentional engagement, in the variable compared to the other conditions. Computational modelling showed that infants' frontal theta power was predicted best by how surprising each movement was. Thus, surprise induced by variability in movements rather than large movements alone engages infants' attention during IDAs. Infants with higher theta power for variable movements were more likely to perform actions successfully and to explore objects novel in the context of the given goal. This highlights the brain mechanisms by which IDAs enhance infants' attention, learning, and exploration.

Habituation Reflects Optimal Exploration Over Noisy Perceptual Samples

From birth, humans constantly make decisions about what to look at and for how long. Yet, the mechanism behind such decision-making remains poorly understood. Here, we present the rational action, noisy choice for habituation (RANCH) model. RANCH is a rational learning model that takes noisy perceptual samples from stimuli and makes sampling decisions based on expected information gain (EIG). The model captures key patterns of looking time documented in developmental research: habituation and dishabituation. We evaluated the model with adult looking time collected from a paradigm analogous to the infant habituation paradigm. We compared RANCH with baseline models (no learning model, no perceptual noise model) and models with alternative linking hypotheses (Surprisal, KL divergence). We showed that (1) learning and perceptual noise are critical assumptions of the model, and (2) Surprisal and KL are good proxies for EIG under the current learning context.

Learning with certainty in childhood

Learners use certainty to guide learning. They maintain existing beliefs when certain, but seek further information when they feel uninformed. Here, we review developmental evidence that this metacognitive strategy does not require reportable processing. Uncertainty prompts nonverbal human infants and nonhuman animals to engage in strategies like seeking help, searching for additional information, or opting out. Certainty directs children's attention and active learning strategies and provides a common metric for comparing and integrating conflicting beliefs across people. We conclude that certainty is a continuous, domain-general signal of belief quality even early in life.

A New Look at Young Children's Referential Informativeness

In this article, I review experimental evidence for the dependence of 2- to 5-year-olds' linguistic referential informativeness on cues to common ground (CG) and propose a process model. Cues to CG provide evidence for CG, that is, for the shared knowledge, beliefs, and attitudes of interlocutors. The presence of cues to CG (e.g., unimpeded listener line of regard or prior mention) is shown to be associated with less informative reference (e.g., pronouns). In contrast, the absence of cues to CG (e.g., impeded listener line of regard or new mention) is shown to be associated with more informative reference (e.g., nouns). Informativeness is sensitive to linguistic before nonlinguistic cues to CG (i.e., 2.0 vs. 2.5 years old, respectively). Reference is cast as a process of active inference, a formulation of Bayesian belief-guided control in biological systems. Child speakers are hierarchical generative models that, characteristically, expect sensory evidence for the evolved, prior Bayesian belief that interlocutor mental states are aligned (i.e., that CG exists). Referential control emerges as an embodied tool to gather evidence for this prior belief. Bottom-up cues to CG elicited by action drive updates to beliefs about CG. In turn, beliefs about CG guide efficient referential control.

The role of mother-infant emotional synchrony in speech processing in 9-month-old infants

Rhythmicity characterizes both interpersonal synchrony and spoken language. Emotions and language are forms of interpersonal communication, which interact with each other throughout development. We investigated whether and how emotional synchrony between mothers and their 9-month-old infants relates to infants' word segmentation as an early marker of language development. Twenty-six 9-month-old infants and their German-speaking mothers took part in the study. To measure emotional synchrony, we coded positive, neutral and negative emotional expressions of the mothers and their infants during a free play session. We then calculated the degree to which the mothers' and their infants' matching emotional expressions followed a predictable pattern. To measure word segmentation, we familiarized infants with auditory text passages and tested how long they looked at the screen while listening to familiar versus novel words. We found that higher levels of predictability (i.e. low entropy) during mother-infant interaction is associated with infants' word segmentation performance. These findings suggest that individual differences in word segmentation relate to the complexity and predictability of emotional expressions during mother-infant interactions.

Studying the Developing Brain in Real-World Contexts: Moving From Castles in the Air to Castles on the Ground

Most current research in cognitive neuroscience uses standardized non-ecological experiments to study the developing brain. But these approaches do a poor job of mimicking the real-world, and thus can only provide a distorted picture of how cognitive operations and brain development unfold outside of the lab. Here we consider future research avenues which may lead to a better appreciation of how developing brains dynamically interact with a complex real-world environment, and how cognition develops over time. We raise several problems faced by current mainstream methods in the field, before briefly reviewing novel promising approaches that alleviate some of these issues. First, we consider research that examines perception by measuring entrainment between brain activity and temporal patterns in naturalistic stimuli. Second, we consider research that examines our ability to parse our continuous experience into discrete events, and how this ability develops over time. Third, we consider the role of children as active agents in selecting what they sample from the environment from one moment to the next. Fourth, we consider new approaches that measure how mutual influences between children and others are instantiated in suprapersonal brain networks. Finally, we discuss how we may reduce adult biases when designing developmental studies. Together, these approaches have great potential to further our understanding of how the developing brain learns to process information, and to control complex real-world behaviors.

Contributions of expected learning progress and perceptual novelty to curiosity-driven exploration

Exploration is curiosity-driven when it relies on the intrinsic motivation to know rather than on extrinsic rewards. Recent evidence shows that artificial agents perform better on a variety of tasks when their learning is curiosity-driven, and humans often engage in curiosity-driven learning when sampling information from the environment. However, which mechanisms underlie curiosity is still unclear. Here, we let participants freely explore different unknown environments that contained learnable sequences of events with varying degrees of noise and volatility. A hierarchical reinforcement learning model captured how participants were learning in these different kinds of unknown environments, and it also tracked the errors they expected to make and the learning opportunities they were planning to seek. With this computational approach, we show that participants' exploratory behavior is guided by learning progress and perceptual novelty. Moreover, we demonstrate an overall tendency of participants to avoid extreme forms of uncertainty. These findings elucidate the cognitive mechanisms that underlie curiosity-driven exploration of unknown environments. Implications of this novel way of quantifying curiosity within a reinforcement learning framework are discussed.

Knowledge in Sight: Toddlers Plan Efficient Epistemic Actions by Anticipating Learning Gains

Anticipating the learning consequences of actions is crucial to plan efficient information seeking. Such a capacity is needed for learners to determine which actions are most likely to result in learning. Here, we tested the early ontogeny of the human capacity to anticipate the amount of learning gained from seeing. In study 1, we tested infants' capacity to anticipate the availability of sight. Fourteen-month-old infants (N = 72) were invited to search for a toy hidden inside a container. The participants were faster to attempt at opening a shutter when this action allowed them to see inside the container. Moreover, this effect was specifically observed when seeing inside the container was potentially useful to the participants' goals. Thus, infants anticipated the availability of sight, and they calibrated their information-seeking behaviors accordingly. In studies 2 and 3, we tested toddlers' capacity to anticipate whether data would be cognitively useful for their goals. Two-and-a-half-year-olds (N = 72) had to locate a target character hidden among distractors. The participants flipped the characters more often, and were comparatively faster to initiate this action when it yielded access to visual data allowing them to locate the target. Thus, toddlers planned their information-seeking behaviors by anticipating the cognitive utility of sight. In contrast, toddlers did not calibrate their behaviors to the cognitive usefulness of auditory data. These results suggest that cognitive models of learning guide toddlers' search for information. The early developmental onset of the capacity to anticipate future learning gains is crucial for active learning.

Co-dependency of exchanged behaviors is a cue for agency attribution in 10-month-olds

Goal-directed social interactions (whether instrumental or communicative) involve co-dependent, partially predictable actions of interacting agents as social goals cannot be achieved by continuously exchanging the same, perfectly predictable, or completely random behaviors. We investigated whether 10-month-olds are sensitive to the co-dependence and degree of predictability in an interactive context where unfamiliar entities exchanged either perfectly predictable (identical), partially predictable (co-dependent), or non-predictable (random) signal sequences. We found that when—following the interactive exchanges—one of the entities turned in the direction of one of two lateral target objects, infants looked more at the indicated referent, but only in the partially predictable signals condition. This shows that infants attributed agency to the orienting entity and interpreted its turning action as a referential object-directed action. The present findings suggest that the co-dependency and partial predictability of exchanged behaviors can serve as an abstract structural cue to attribute intentional agency and recognize goal-directed social interactions.