Sepehr Mahmoudian

Stealth AI startup.

Responsibilities:• Working with a team with a team of linguistic experts to do basic research on human language and cognition with a highly advanced multi-area model of the brain called MatCo12.• Software engineering to scale up models, optimization to increase performance, unit testing, CI/CL in C/C++ and Python.• Large-scale simulations of recurrent and hierarchical neural networks on parallelized and distributed architectures with OpenMP and MPI.• Statistical analysis of data using machine learning techn

Responsibilities:• Creating functional brain-inspired deep hierarchical neural networks In the Wibral lab Göttingen.• Using the new mathematics of partial information decomposition to create novel context-aware neurons in C/C++ with unique objective functions specified in extended information theoretic terms.• Optimizing and running large-scale simulations of neural networks in parallel and distributed modes.• Analysis of simulation results in Python with information theory and machine learning techniques.•

Functional novel partial information cognitive architectures comprised of large-scale deep hierarchical neural networks informed and constrained by experimental neuroanatomical and neurophysiological data. Using statistical machine learning and artificial intelligence techniques to analyze the outp