Learning

A persistent puzzle throughout the cognitive and social sciences is how people manage to learn social conventions from the sparse and noisy behavioral data of diverse actors, without explicit instruction. Here, we show that the dominant theories of social learning perform poorly at capturing how individuals learn conventions in coordination experiments that task them with matching their behaviors while interacting in social networks. Across experiments, participants’ choices systematically deviate from both imitation and optimization. Instead, they follow a categorical, two-stage learning (...) process: they behave probabilistically until they acquire enough information about each other to trigger a mental threshold and then their choices stabilize. We effectively estimate this threshold using the tolerance principle (TP), a parameter-free equation developed to model how children learn rules in language. We show that threshold-based agents produce social learning that is more accurate than imitating and optimizing agents, while also providing a better model of how a critical mass of dissenters can overturn conventions. The superior performance of our model holds when comparing against a variety of optimization approaches, including Bayesian inference. Furthermore, in a preregistered dyadic experiment requiring people to infer nonlinguistic behavioral patterns amid controlled levels of noise in observed signals, TP outperforms all other models at reproducing learning rates among human participants. These findings offer compelling evidence that a simple, mathematical threshold underlies individual and social learning, from grammatical rules to behavioral conventions. (shrink)

What is the difference between rational forms of learning and ‘arational’ forms of attitude change? I argue that rational learning occurs when a change in attitude is based on the acquisition of new information. I then attempt to state precisely what this requirement amounts to.

The geometry of the environment can affect numerous psychological, social, and ecological processes. But its roles in social learning and the dynamics of descriptive norms remain unclear. Here we use agent-based modeling to explore how environments with different geometric shapes can constrain social learning to produce universally shared descriptive norms. Our simulations show that an environment with an irregular layout facilitates the emergence of multiple descriptive norms in a population, whereas an environment with a regular grid plan constrains social learning (...) to produce a behaviorally homogeneous population. (shrink)

Regret compounds the pain of mistakes by making us ruminate on the past. Is there any value in this compounded suffering? An intuitive and widely endorsed view, the learning view, states that regret has value because it helps us learn from our mistakes. This paper challenges the learning view. I show that we often learn from our mistakes without regret, through a process I call “belief‐driven learning.” The challenge is to explain what value regret adds to our lives, over and (...) above belief‐driven learning. To meet the challenge, I develop the idea that regret helps us overcome temptation. I show that belief‐driven learning falters for mistakes made due to temptation because the source of error is not epistemic, but motivational. I argue that the ruminative processes involved in regret can help us learn in the face of temptation by reshaping our motivational tendencies, bringing them into better alignment with the values we endorse. (shrink)

Note: I am currently rewriting this paper, having learned a lot since I wrote this version, but this version might still be of interest and I have also cited it in another paper, so I am archiving it here.

Despite excelling in high-level reasoning, current language models lack robustness in real-world scenarios and perform poorly on fundamental problem-solving tasks that are intuitive to humans. This paper argues that both challenges stem from a core discrepancy between human and machine cognitive development. While both systems rely on increasing representational power, the absence of core knowledge-foundational cognitive structures in humans-prevents language models from developing robust, generalizable abilities, where complex skills are grounded in simpler ones within their respective domains. It explores empirical (...) evidence of core knowledge in humans, analyzes why language models fail to acquire it, and argues that this limitation is not an inherent architectural constraint. Finally, it outlines a workable proposal for systematically integrating core knowledge into future multi-modal language models through the large-scale generation of synthetic training data using a cognitive prototyping strategy. (shrink)

The origin of memories is thought to be found in sensory perception. This conception is central to how the memory sciences characterize encoding. This paper considers how novel memory traces can be formed independently of external sensory inputs. I present a case study in which memory traces are created without sensory perception using a technique I call optogenetic memory implantation. Comparing this artificial process with normal memory encoding, I consider its implications for rethinking the causal chain that leads to remembering.

The notions of psychological similarity and probabilistic learning are key posits in cognitive, computational, and developmental psychology and in machine learning. However, their explanatory relationship is rarely made explicit within and across these research fields. This opinionated review critically evaluates how these notions can mutually inform each other within computational cognitive science. Using probabilistic models of concept learning as a case study, I argue that two notions of psychological similarity offer important normative constraints to guide modelers’ interpretations of representational primitives. (...) In particular, the two notions furnish probabilistic models of cognition with meaningful interpretations of what the associated subjective probabilities in the model represent and how they attach to experiences from which the agent learns. Similarity representations thereby provide probabilistic models with cognitive, as opposed to purely mathematical, content. (shrink)

Formal representations drawn from rational choice theory have been used in a variety of ways to fruitfully model the way in which actual agents are approximately rational. This analysis requires bridging between ideal normative theory, in which the mechanisms, representations, and other such internal parts are in an important sense interchangeable, and descriptive psychological theory, in which understanding the internal workings of the agent is often the main goal of the entire inquiry. In this paper, I raise a problem brought (...) on by this gap: for almost every theory of approximate rationality, there will be an empirically indistinguishable alternative that individuates the parts of the agent in a significantly different way. (shrink)

The aim of this paper is to understand the functional role of mental representations and intentionality in skilled actions from a systems related perspective. Therefore, we will evaluate the function of representation and then discuss the cognitive architecture of skilled actions in more depth. We are going to describe the building blocks and levels of the action system that enable us to control movements such as striking the tennis ball at the right time, or grasping tools in manual action. Based (...) on this theoretical understanding the measurement of mental representations and related research results concerning mental representation in skilled action are presented in an overview. This leads to the question how mental representations develop and change during learning. Finally, to consolidate the functional understanding of mental representation in skilled action and interaction, we provide examples how to use the measurement of mental representation in humans to inform technical systems. (shrink)

Learning in educational settings most often emphasizes declarative and procedural knowledge. Studies of expertise, however, point to other, equally important components of learning, especially improvements produced by experience in the extraction of information: Perceptual learning. Here we describe research that combines principles of perceptual learning with computer technology to address persistent difficulties in mathematics learning. We report three experiments in which we developed and tested perceptual learning modules to address issues of structure extraction and fluency in relation to algebra and (...) fractions. PLMs focus students’ learning on recognizing and discriminating, or mapping key structures across different representations or transformations. Results showed significant and persisting learning gains for students using PLMs. PLM technology offers promise for addressing neglected components of learning: Pattern recognition, structural intuition, and fluency. Using PLMs as a complement to other modes of instruction may allow students to overcome chronic problems in learning. (shrink)

Heyes’ book is an essential addition to the literature on human uniqueness. Her main claim is that the key human cognitive capacities are products of cultural rather than genetic evolution. Among these distinctively human capacities are causal understanding, episodic memory, imitation, mindreading, and normative thinking. According to Heyes, they emerged not by genetic mutation but by innovations in cognitive development. She calls these mechanisms ‘cognitive gadgets.’ This is perhaps one of the best and most comprehensive views of human cognitive evolution (...) advanced in recent years. (shrink)

I hypothesize that re-occurring prior experience of complex systems mobilizes a fast response, whose attractor is encoded by their strongly connected network core. In contrast, responses to novel stimuli are often slow and require the weakly connected network periphery. Upon repeated stimulus, peripheral network nodes remodel the network core that encodes the attractor of the new response. This “core-periphery learning” theory reviews and generalizes the heretofore fragmented knowledge on attractor formation by neural networks, periphery-driven innovation, and a number of recent (...) reports on the adaptation of protein, neuronal, and social networks. The core-periphery learning theory may increase our understanding of signaling, memory formation, information encoding and decision-making processes. Moreover, the power of network periphery-related “wisdom of crowds” inventing creative, novel responses indicates that deliberative democracy is a slow yet efficient learning strategy developed as the success of a billion-year evolution. Also see the video abstract here: https://youtu.be/IIjP7zWGjVE. The network core triggers fast responses to previously experienced stimuli. Innovations require the network periphery that encodes a novel attractor by core-remodeling. This core-periphery learning theory is introduced as a general adaptation, learning, decision-making, and forgetting mechanism of all complex systems including protein, metabolic, signaling, neuronal, social networks, and ecosystems. (shrink)

This paper considers the nature of learning and the role of institutions in general and schools in particular in structuring learning. It outlines and commends a view of learning as a process whereby we discover ourselves as persons and thereby act to create the contexts in which we live and work. Central to this view is the idea of the ‘learning school’.

A thorough survey of the field of learning. _Learning & Behavior_ covers topics such as classical and operant conditioning, reinforcement schedules, avoidance and punishment, stimulus control, comparative cognition, observational learning, motor skill learning, and choice. The book includes thorough coverage of classic studies and the most recent developments and trends, while providing examples of real-world applications of the principles discovered in laboratory research. It also emphasizes the behavioral approach but not exclusively so; many cognitive theories are covered as well, and (...) there is a chapter on comparative cognition. Learning Goals Upon completing this book readers will be able to: Understand the field of learning Discuss real-world applications of learning principles. (shrink)

Defines learning and shows how the learning process is studied. Clearly written and user-friendly, _Introduction to the Theories of Learning_ places learning in its historical perspective and provides appreciation for the figures and theories that have shaped 100 years of learning theory research. The 9th edition has been updated with the most current research in the field. Learning Goals Upon completing this book, readers should be able to: Define learning and show how the learning process is studied Place learning theory (...) in historical perspective Present essential features of the major theories of learning with implications for educational practice. (shrink)

Autonomous development of gaze behavior for natural human-robot interaction.

Most of the existing classification techniques concentrate on learning the datasets as a single similar unit, in spite of so many differentiating attributes and complexities involved. However, traditional classification techniques are required to analyze the datasets prior to learning, and if not doing so, they loss their performance in terms of accuracy and AUC. To this end, many of the machine learning problems can be very easily solved just by carefully observing human learning and training nature and then mimicking the (...) same in the machine learning. In response to these issues, we present a comprehensive suite of experiments carefully designed to provide conclusive, reliable, and significant results to the problem of efficient learning. This paper proposes a novel, simple, and effective machine learning paradigm that explicitly exploits this important similar-to-different (S2D) human learning strategy and implements it based on two algorithms (C4.5 and CART) efficiently. The framework not only analyzes the data sets prior to implementation, but also carefully allows classifier to have a systematic study so as to mimic the human training technique designed for efficient learning. Experimental results show that the method outperforms the state-of-the-art methods in terms of learning capability and breaks through the gap between human and machine learning. In fact, the proposed similar-to-different (S2D) strategy may also be useful in efficient learning of real-world complex and high-dimensional data sets, especially which are very typical to learn with traditional classifiers. (shrink)

Rather than simply outlining the classical and modern theories of learning, this widely adopted text brings the material to life through case studies that engage students in debates about what really happens in classrooms. Students are encouraged to test the strengths and weaknesses of each theory so that, ultimately, they will learn to formulate their own philosophies of teaching and learning. The new Fifth Edition of Perspectives on Learning features: A discussion of common sense and learning theories. A new chapter (...) on Transfer of Learning. Consideration of recent developments in brain science. A thoroughly updated list of Recommendations for Further Reading. Perspectives on Learning is one of the five books in the highly regarded Teachers College Press THINKING ABOUT EDUCATION SERIES, now in its Fifth Edition. All of the books in this series are designed to help pre- and in-service teachers bridge the gap between theory and practice. (shrink)

The mind's basic task is to organize adaptive behaviour. I argue that necessary conditions to achieve this are acquiring a 'body-self', a differentiated perception, motor intuition, and motor control. The latter three can be learnied implicitly by crosswise comparing the perceived actual situation, the desired situation, the perceived result and the anticipated result.