40.7128° N, 74.0060° W
Cortical Mapping System v0.4.2
Frontier AI Researcher & Engineer

INIT ARCHITECTURE:Arash Nikzad

"Synthesizing Biological Intelligence with Machine Architectures"

I specialize in the intersection of computational neuroscience and deep learning, building autonomous systems inspired by the Thousand Brains Theory to achieve robust, hierarchical world modeling.

SEC_ABOUT

About Me

AN

Arash Nikzad

RESEARCH_IDENTIFIER: 8842-X

FOCUS:CORTICAL_ARCH
LOCATION:DISTRIBUTED
STATUS:ACTIVE_RESEARCH
[ START_BIO_TRANSMISSION ]

My research trajectory is centered on the formalization of neocortical principles into scalable machine learning frameworks. Deeply influenced by the Thousand Brains Theory of Intelligence, I develop systems that transition from static pattern recognition to dynamic, sensory-motor world modeling.

Currently, I am investigating how sparse distributed representations, grid-cell based navigation, and hierarchical predictive coding can be integrated into modern RL and generative agents. My goal is to build intelligence that doesn't just process data, but understands the underlying structural causal models of its environment.

[ END_BIO_TRANSMISSION ]
SEC_SKILLS

Technical Skills

MODULE_1

Theoretical AI/ML

Deep LearningActive InferenceRepresentation LearningGenerative AIProbabilistic Graph Models
MODULE_2

Computational Neuroscience

Cortical Column ModelingSDR TheorySequence MemoryGrid Cell NavigationCognitive Architectures
MODULE_3

Core Engineering

PythonC++ / CUDASystem ArchitectureHigh Performance ComputingLinux Kernel
MODULE_4

Research Frameworks

PyTorchJAXNumenta NuPICTensorFlowWeights & Biases
MODULE_5

Mathematical Foundations

Stochastic CalculusInformation GeometryDifferential GeometryNeural Dynamics
MODULE_6

Deployment Systems

DockerKubernetesAWS Research CloudNVIDIA TritonMLOps
SEC_PROJECTS

Featured Projects

PROJECT_TYPE: Core ResearchREF_ID: 001-SR

Cortical-Net v1.0

A large-scale implementation of a thousand-brains inspired architecture, achieving superior generalization in 3D object recognition through sensory-motor integration.

PyTorchCUDANeuroSim
STATUS: VERIFIEDVIEW_DATA
PROJECT_TYPE: Generative AIREF_ID: 002-SR

VQ-VAE + HMM Sequence Engine

A hybrid probabilistic-generative model for unsupervised discovery of temporal hierarchies in high-dimensional video streams.

JAXHaikuProbabilistic Models
STATUS: VERIFIEDVIEW_DATA
PROJECT_TYPE: Robotics/RLREF_ID: 003-SR

Active-Sensation RL Agent

Autonomous agent utilizing grid-cell representations for efficient exploration and mapping of non-Euclidean environments.

PyTorchGymnasiumStable Baselines3
STATUS: VERIFIEDVIEW_DATA
PROJECT_TYPE: Systems EngineeringREF_ID: 004-SR

Distributed Neocortical Simulator

High-performance C++ simulator for studying the synchronization and voting dynamics between thousands of simulated cortical columns.

C++20MPIOpenMP
STATUS: VERIFIEDVIEW_DATA
PROJECT_TYPE: Applied AIREF_ID: 005-SR

LLM Travel Reasoning Workflow

Multi-agent system that leverages hierarchical planning and constraint satisfaction for complex, long-horizon travel optimization.

LangChainOpenAIPython
STATUS: VERIFIEDVIEW_DATA
SEC_RESEARCH

Research & Vision

Developing a unified framework for intelligence based on the distributed, voting-based architecture of the neocortex.

RE_STREAM_1

Cortical Micro-Architectures

Modeling individual cortical columns as independent, sensory-motor learning units capable of object-centric representation.

RE_STREAM_2

SDR & Memory Systems

Utilizing Sparse Distributed Representations (SDRs) to achieve high-capacity, robust, and biologically plausible memory.

RE_STREAM_3

Hierarchical World Models

Building multi-level predictive systems that capture environmental regularities through active inference and grid cell mapping.

RE_STREAM_4

Biomimetic RL Agents

Engineering agents that utilize active sensation and internal coordinate systems for complex navigation and manipulation.

SEC_EXPERIENCE

Professional Journey

01

Lead Research Engineer

2023 - PRESENT

Frontier Neuroscience Lab

Architecting large-scale cortical models and optimizing neural simulation kernels for distributed GPU clusters.

02

Graduate Researcher

2021 - 2023

Advanced Intelligence Institute

Developed hierarchical representation learning algorithms and contributed to the peer-reviewed research on sensory-motor integration.

SEC_CONTACT

Get In Touch

[ ESTABLISH_COMMUNICATION ]

Ready for Collaboration

I am actively seeking research collaborations, technical discussions, and opportunities at the frontier of AI and neuroscience.