Geodynamic zoning of the southwestern part of the Talnakh Orogenic System

This study focuses on the markers of tectonically stressed zones inside the rock mass, that were identified during the regional geodynamic zoning of the mine fields of the Talnakh orogenic system. Identification features for tracing geodynamically active structures within the western flank of the Talnakh orogenic system have been identified based on morphometric analysis of the Tunguska series sediments, which are the upper layer of the ore-bearing intrusions and associated ore deposits. In the larger morphostructural groups, the boundaries of contrastingly alternating zones of elevated and depressed absolute depths at the base and the roof of the Tunguska series sediments represent the boundaries of tectonic blocks of different elevation levels with sharply contrasting indices of terrain stress. The circular-shaped structures highlighted in the morphostructural schemes spatially coincide with the tectonic forms were formed as the result of strike-slip and torsional processes. A heterogeneity, which is reflected in the allocation of blocks with different values of the stress distribution coefficient (K) is identified in the initial stress field of the Tunguska series sediments. The boundaries of the geodynamic blocks that were identified using to different methods are identical. It is established that the assumed faults correspond to the faults identified based on the detailed exploration data.

Efficient coding of natural scenes improves neural system identification

Neural system identification aims at learning the response function of neurons to arbitrary stimuli using experimentally recorded data, but typically does not leverage coding principles such as efficient coding of natural environments. Visual systems, however, have evolved to efficiently process input from the natural environment. Here, we present a normative network regularization for system identification models by incorporating, as a regularizer, the efficient coding hypothesis, which states that neural response properties of sensory representations are strongly shaped by the need to preserve most of the stimulus information with limited resources. Using this approach, we explored if a system identification model can be improved by sharing its convolutional filters with those of an autoencoder which aims to efficiently encode natural stimuli. To this end, we built a hybrid model to predict the responses of retinal neurons to noise stimuli. This approach did not only yield a higher performance than the stand-alone system identification model, it also produced more biologically-plausible filters. We found these results to be consistent for retinal responses to different stimuli and across model architectures. Moreover, our normatively regularized model performed particularly well in predicting responses of direction-of-motion sensitive retinal neurons. In summary, our results support the hypothesis that efficiently encoding environmental inputs can improve system identification models of early visual processing.