University of St Andrews

About: University of St Andrews is a education organization based out in St Andrews, Fife, United Kingdom. It is known for research contribution in the topics: Population & Laser. The organization has 16260 authors who have published 43364 publications receiving 1636072 citations. The organization is also known as: St Andrews University & University of St. Andrews.

A selective impairment in the processing of sad and fearful expressions in children with psychopathic tendencies.

Abstract: The processing of emotional expressions is fundamental for normal socialisation and interaction. Reduced responsiveness to the expressions of sadness and fear has been implicated in the development of psychopathy (R. J. R. Blair, 1995). The current study investigates the sensitivity of children with psychopathic tendencies to facial expressions. Children with psychopathic tendencies and a comparison group, as defined by the Psychopathy Screening Device (PSD; P. J. Frick & R. D. Hare, in press), were presented with a cinematic display of a standardised set of facial expressions that depicted sadness, happiness, anger, disgust, fear, and surprise. Participants observed as these facial expressions slowly evolved through 20 successive frames of increasing intensity. The children with psychopathic tendencies presented with selective impairments; they needed significantly more stages before they could successfully recognise the sad expressions and even when the fearful expressions were at full intensity were significantly more likely to mistake them for another expression. These results are interpreted with reference to an amygdala and empathy impairment explanation of psychopathy.

How culture shaped the human genome: bringing genetics and the human sciences together

Abstract: Researchers from diverse backgrounds are converging on the view that human evolution has been shaped by gene-culture interactions. Theoretical biologists have used population genetic models to demonstrate that cultural processes can have a profound effect on human evolution, and anthropologists are investigating cultural practices that modify current selection. These findings are supported by recent analyses of human genetic variation, which reveal that hundreds of genes have been subject to recent positive selection, often in response to human activities. Here, we collate these data, highlighting the considerable potential for cross-disciplinary exchange to provide novel insights into how culture has shaped the human genome.

A Change in the Geodynamics of Continental Growth 3 Billion Years Ago

Abstract: Models for the growth of continental crust rely on knowing the balance between the generation of new crust and the reworking of old crust throughout Earth’s history. The oxygen isotopic composition of zircons, for which uranium-lead and hafnium isotopic data provide age constraints, is a key archive of crustal reworking. We identified systematic variations in hafnium and oxygen isotopes in zircons of different ages that reveal the relative proportions of reworked crust and of new crust through time. Growth of continental crust appears to have been a continuous process, albeit at variable rates. A marked decrease in the rate of crustal growth at ~3 billion years ago may be linked to the onset of subduction-driven plate tectonics.

Random synaptic feedback weights support error backpropagation for deep learning.

Abstract: The brain processes information through multiple layers of neurons. This deep architecture is representationally powerful, but complicates learning because it is difficult to identify the responsible neurons when a mistake is made. In machine learning, the backpropagation algorithm assigns blame by multiplying error signals with all the synaptic weights on each neuron’s axon and further downstream. However, this involves a precise, symmetric backward connectivity pattern, which is thought to be impossible in the brain. Here we demonstrate that this strong architectural constraint is not required for effective error propagation. We present a surprisingly simple mechanism that assigns blame by multiplying errors by even random synaptic weights. This mechanism can transmit teaching signals across multiple layers of neurons and performs as effectively as backpropagation on a variety of tasks. Our results help reopen questions about how the brain could use error signals and dispel long-held assumptions about algorithmic constraints on learning. Multi-layered neural architectures that implement learning require elaborate mechanisms for symmetric backpropagation of errors that are biologically implausible. Here the authors propose a simple resolution to this problem of blame assignment that works even with feedback using random synaptic weights.