Scientific racism | EPFL Graph Search

Scientific racism, sometimes termed biological racism, is the pseudoscientific belief that the human species can be subdivided into biologically distinct taxa called "races", and that empirical evidence exists to support or justify racism (racial discrimination), racial inferiority, or racial superiority. Before the mid-20th century, scientific racism received credence throughout the scientific community, but it is no longer considered scientific. The division of humankind into biologically distinct groups, and the attribution of specific traits both physical and mental to them by constructing and applying corresponding explanatory models, that is, racial theories, is sometimes called racialism, race realism, or race science by its proponents. Modern scientific consensus rejects this view as being irreconcilable with modern genetic research. Scientific racism misapplies, misconstrues, or distorts anthropology (notably physical anthropo

Biologically plausible unsupervised learning in shallow and deep neural networks

Bernd Albert Illing

The way our brain learns to disentangle complex signals into unambiguous concepts is fascinating but remains largely unknown. There is evidence, however, that hierarchical neural representations play a key role in the cortex. This thesis investigates biologically plausible models of unsupervised learning of hierarchical representations as found in the brain and modern computer vision models. We use computational modeling to address three main questions at the intersection of artificial intelligence (AI) and computational neuroscience.The first question is: What are useful neural representations and when are deep hierarchical representations needed? We approach this point with a systematic study of biologically plausible unsupervised feature learning in a shallow 2-layer networks on digit (MNIST) and object (CIFAR10) classification. Surprisingly, random features support high performance, especially for large hidden layers. When combined with localized receptive fields, random feature networks approach the performance of supervised backpropagation on MNIST, but not on CIFAR10. We suggest that future models of biologically plausible learning should outperform such random feature benchmarks on MNIST, or that such models should be evaluated in different ways.The second question is: How can hierarchical representations be learned with mechanisms supported by neuroscientific evidence? We cover this question by proposing a unifying Hebbian model, inspired by common models of V1 simple and complex cells based on unsupervised sparse coding and temporal invariance learning. In shallow 2-layer networks, our model reproduces learning of simple and complex cell receptive fields, as found in V1. In deeper networks, we stack multiple layers of Hebbian learning but find that it does not yield hierarchical representations of increasing usefulness. From this, we hypothesise that standard Hebbian rules are too constrained to build increasingly useful representations, as observed in higher areas of the visual cortex or deep artificial neural networks.The third question is: Can AI inspire learning models that build deep representations and are still biologically plausible? We address this question by proposing a learning rule that takes inspiration from neuroscience and recent advances in self-supervised deep learning. The proposed rule is Hebbian, i.e. only depends on pre- and post-synaptic neuronal activity, but includes additional local factors, namely predictive dendritic input and widely broadcasted modulation factors. Algorithmically, this rule applies self-supervised contrastive predictive learning to a causal, biological setting using saccades. We find that networks trained with this generalised Hebbian rule build deep hierarchical representations of images, speech and video.We see our modeling as a potential starting point for both, new hypotheses, that can be tested experimentally, and novel AI models that could benefit from added biological realism.

EPFL2021

Does Stanford's induction apply to engineering sciences?

Tao-Chun Lee

Recent works in scientific realism by Stanford (2006) and Chakravartty (2008) have generated interests in the research of a more selective and sophisticated scientific realism. Such debates advance the philosophical development of scientific realism, and arguably shape scientific realism toward a more refined description. Among many scientific theories discussed by Stanford (2006) and Chakravartty (2008), we have found their examples excluding those from engineering sciences, and this attracts our attention. In this paper, we first give an overview of Stanford's induction. Second, we analyze the arguments presented by Stanford (2006) and Chakravartty (2008) in depth to explore possible reasons why engineering sciences are seemly out of the scope of the debates. Third, we re-examine the philosophical debates by Stanford (2006) and Chakravartty (2008) in the context of an interesting case in mechanical engineering, d'Alembert's paradox, presented by Grimberg (2008). We conclude in the final section on the validity of applying Stanford's induction to engineering sciences.

2012

Minimalist Design of a 3-Axis Passive Compliant Foot for Sprawling Posture Robots

Tomislav Horvat, Auke Ijspeert, Kamilo Andres Melo Becerra

A sprawled posture amphibious biorobot resembling a salamander had helped us in conquering scientific questions of the locomotion of these animals and offer us with technological possibilities for applications in disaster response. However, so far the foot structure of these robots is simplified in a ball foot that occludes many interesting aspects of the rich dynamics of interaction with the ground that these animals/robots have. In this paper, we present a minimalist design of a three degree of freedom foot that uses passive mechanics, soft materials and simple fabrication techniques to achieve features observed in sprawling posture animals walking gaits. The parameters of our design can be adjusted for specific visco-elastic properties (stiffness, damping) and easily adapted to different sizes and forms. We presented the fabrication technique used to achieve several configurations of the foot structure. Stress analysis of the design helped us to verify the right selection of materials and configurations to achieve desired behaviors. We expect to use these feet for better understanding the limb-ground interaction in sprawled animals as well as improving the locomotion capabilities of biorobots in general.

IEEE2019

Biologically plausible unsupervised learning in shallow and deep neural networks

Bernd Albert Illing

EPFL2021