Mind the gap
Related Articles
Spontaneous thought separates into clusters of negative, positive, and flexible thinking
The nature and frequency of spontaneous thoughts play a critical role in cognitive processes like perception, decision-making, attention, and memory. Deficits in these processes are also greatly associated with the development and maintenance of psychopathology. However, the underlying cognitive dynamics of free and stuck spontaneous thought remain unclear, as these often occur in the absence of measurable behaviors. Here, we analyze free word-association data using attractor-state dynamic modeling, which conceptualizes stuck spontaneous thought as navigating a multidimensional semantic space while in the presence of strong attractor locations. Word-association data was collected from an exploratory sample (N1 = 65), a first replication sample (N2 = 79), and, following pre-registration, a second replication sample (N3 = 222). After the data was embedded into a 3-dimensional semantic space and fit by our dynamic model, unsupervised learning consistently grouped data into four clusters across all independent samples. These clusters were characterized by two distinct patterns of stuck negative thinking, a pattern of protective positive thinking, and a pattern of flexible mind-wandering. Our results support a method for modeling spontaneous thought and isolate distinct sub-types that may not be accessible using retrospective self-report methods. We discuss implications for clinical and cognitive science.
The individual determinants of morning dream recall
Evidence suggests that (almost) everyone dreams during their sleep and may actually do so for a large part of the night. Yet, dream recall shows large interindividual variability. Understanding the factors that influence dream recall is crucial for advancing our knowledge regarding dreams’ origin, significance, and functions. Here, we tackled this issue by prospectively collecting dream reports along with demographic information and psychometric, cognitive, actigraphic, and electroencephalographic measures in 217 healthy adults (18–70 y, 116 female participants, 101 male participants). We found that attitude towards dreaming, proneness to mind wandering, and sleep patterns are associated with the probability of reporting a dream upon morning awakening. The likelihood of recalling dream content was predicted by age and vulnerability to interference. Moreover, dream recall appeared to be influenced by night-by-night changes in sleep patterns and showed seasonal fluctuations. Our results provide an account for previous observations regarding inter- and intra-individual variability in morning dream recall.
An integrative data-driven model simulating C. elegans brain, body and environment interactions
The behavior of an organism is influenced by the complex interplay between its brain, body and environment. Existing data-driven models focus on either the brain or the body–environment. Here we present BAAIWorm, an integrative data-driven model of Caenorhabditis elegans, which consists of two submodels: the brain model and the body–environment model. The brain model was built by multicompartment models with realistic morphology, connectome and neural population dynamics based on experimental data. Simultaneously, the body–environment model used a lifelike body and a three-dimensional physical environment. Through the closed-loop interaction between the two submodels, BAAIWorm reproduced the realistic zigzag movement toward attractors observed in C. elegans. Leveraging this model, we investigated the impact of neural system structure on both neural activities and behaviors. Consequently, BAAIWorm can enhance our understanding of how the brain controls the body to interact with its surrounding environment.
Experimental observation of gapped shear waves and liquid-like to gas-like dynamical crossover in active granular matter
Unlike crystalline solids, liquids lack long-range order, resulting in diffusive shear fluctuations rather than propagating waves. Simulations predict that liquids exhibit a k-gap in wave-vector space, where solid-like transverse waves reappear above this gap. Experimental evidence in classical liquids has been limited, observed only in 2D dusty plasmas. Here, we investigate this phenomenon using active Brownian vibrators and uncover distinct gas-like and liquid-like phases depending on the packing fraction. We measure key properties, including pair correlation functions, mean square displacements, velocity auto-correlation functions, and vibrational density of states. In the liquid-like phase, we confirm the k-gap in transverse excitations, whose size grows as the packing fraction decreases and eventually disappears in the gas phase. Our findings extend the concept of the k-gap to active granular systems and reveal striking parallels with supercritical fluids.
What large language models know and what people think they know
As artificial intelligence systems, particularly large language models (LLMs), become increasingly integrated into decision-making processes, the ability to trust their outputs is crucial. To earn human trust, LLMs must be well calibrated such that they can accurately assess and communicate the likelihood of their predictions being correct. Whereas recent work has focused on LLMs’ internal confidence, less is understood about how effectively they convey uncertainty to users. Here we explore the calibration gap, which refers to the difference between human confidence in LLM-generated answers and the models’ actual confidence, and the discrimination gap, which reflects how well humans and models can distinguish between correct and incorrect answers. Our experiments with multiple-choice and short-answer questions reveal that users tend to overestimate the accuracy of LLM responses when provided with default explanations. Moreover, longer explanations increased user confidence, even when the extra length did not improve answer accuracy. By adjusting LLM explanations to better reflect the models’ internal confidence, both the calibration gap and the discrimination gap narrowed, significantly improving user perception of LLM accuracy. These findings underscore the importance of accurate uncertainty communication and highlight the effect of explanation length in influencing user trust in artificial-intelligence-assisted decision-making environments.
Responses