scholarly journals Parametric Control of Flexible Timing Through Low-Dimensional Neural Manifolds

2021 ◽  
Author(s):  
Manuel Beiran ◽  
Nicolas Meirhaeghe ◽  
Hansem Sohn ◽  
Mehrdad Jazayeri ◽  
Srdjan Ostojic
Keyword(s):  
Parametric Control ◽  
Low Dimensional

scholarly journals Parametric control of flexible timing through low-dimensional neural manifolds

2021 ◽  
Author(s):  
Manuel Beiran ◽  
Nicolas Meirhaeghe ◽  
Hansem Sohn ◽  
Mehrdad Jazayeri ◽  
Srdjan Ostojic
Keyword(s):  
Recurrent Neural Networks ◽  
Behavioral Responses ◽  
Dimensional Subspace ◽  
Neural Data ◽  
Parametric Control ◽  
Neural Processes ◽  
Low Dimensional ◽  
Behaving Monkeys ◽  
Open Question ◽  
Theoretical Analyses

Biological brains possess an unparalleled ability to generalize adaptive behavioral responses from only a few examples. How neural processes enable this capacity to extrapolate is a fundamental open question. A prominent but underexplored hypothesis suggests that generalization is facilitated by a low-dimensional organization of collective neural activity. Here we tested this hypothesis in the framework of flexible timing tasks where dynamics play a key role. Examining trained recurrent neural networks we found that confining the dynamics to a low-dimensional subspace allowed tonic inputs to parametrically control the overall input-output transform and enabled smooth extrapolation to inputs well beyond the training range. Reverse-engineering and theoretical analyses demonstrated that this parametric control of extrapolation relies on a mechanism where tonic inputs modulate the dynamics along non-linear manifolds in activity space while preserving their geometry. Comparisons with neural data from behaving monkeys confirmed the geometric and dynamical signatures of this mechanism.

Thermoelectric Properties of PbSr(Se,Te)-Based Low Dimensional Structures

2000 ◽  
Vol 626 ◽  
Author(s):  
Harald Beyer ◽  
Joachim Nurnus ◽  
Harald Böttner ◽  
Armin Lambrecht ◽  
Lothar Schmitt ◽  
...  
Keyword(s):  
Thermal Conductivity ◽  
Thermoelectric Properties ◽  
Power Factor ◽  
Evaluation Method ◽  
Interband Transition ◽  
Multiple Quantum Well ◽  
Calculated Data ◽  
Multiple Quantum ◽  
Ir Transmission ◽  
Low Dimensional

ABSTRACTThermoelectric properties of low dimensional structures based on PbTe/PbSrTe-multiple quantum-well (MQW)-structures with regard to the structural dimensions, doping profiles and levels are presented. Interband transition energies and barrier band-gap are determined from IR-transmission spectra and compared with Kronig-Penney calculations. The influence of the data evaluation method to obtain the 2D power factor will be discussed. The thermoelectrical data of our layers show a more modest enhancement in the power factor σS2 compared with former publications and are in good agreement with calculated data from Broido et al. [5]. The maximum allowed doping level for modulation doped MQW structures is determined. Thermal conductivity measurements show that a ZT enhancement can be achieved by reducing the thermal conductivity due to interface scattering. Additionally promising lead chalcogenide based superlattices for an increased 3D figure of merit are presented.

Sign in / Sign up

Export Citation Format

Share Document