Abrhame Habtom
Computer Science Student & AI Enthusiast
Computer Science student at Addis Ababa University with a strong interest in AI and machine learning. Currently interning at iCog Labs, exploring neural-symbolic AI integration and working on cognitive architecture projects.
About Me
A passionate computer science student with a growing interest in AI and machine learning, eager to contribute to the advancement of artificial intelligence through research and development.
I am currently pursuing my Bachelor's degree in Computer Science at Addis Ababa University, with an expected graduation in July 2027. My academic journey has been complemented by hands-on experience as an intern at iCog Labs, where I work on AI research projects.
My interests lie at the intersection of symbolic reasoning and neural networks. I'm fascinated by how AI systems can combine pattern recognition with logical reasoning to create more explainable and trustworthy artificial intelligence.
Through my internship at iCog Labs, I've been learning about MeTTa programming, OpenPSI cognitive architecture, and various machine learning frameworks. I'm particularly interested in exploring how reinforcement learning can be integrated into neural-symbolic systems.
Education
BSc Computer Science
Addis Ababa University
Location
Addis Ababa, Ethiopia
Current Role
AI Research Intern
iCog Labs
Featured Projects
A selection of my best work in machine learning, AI research, and software development.
Contributing to MeTTa experiments for the OpenPSI cognitive architecture. Working on neural-symbolic AI integration and cognitive modeling systems.
Implementation of UNet architecture for diffusion models, exploring generative AI and deep learning techniques for image generation.
Implementation of federated learning algorithms, exploring distributed machine learning approaches for privacy-preserving AI systems.
Comprehensive collection of machine learning and AI projects built with PyTorch, including model building, training, and testing implementations.
From-scratch implementation of neural networks, demonstrating understanding of fundamental deep learning concepts and algorithms.
Experience at iCog Labs
My journey as an AI research intern, working on cognitive architecture development and neural-symbolic integration projects.
Implementing the OpenPSI codebase using MetTaLog, adapting functions that don't work in MetTaLog environment.
Integrated Mersenne Twister random number generator into MetTaLog language and OpenPSI codebase.
Built a graphical visualizer using matplotlib to showcase the emotions of the curious agent in real-time.
Refactored the entire codebase to increase the use of built-in functions like map-atom and fold1-atom instead of utility module implementations.
Developed MeTTa rule expression to Python schema object converter for agent emotion processing.
Implemented and completed the ping-pong use case experiment in MeTTa for cognitive agent testing.
Developed and implemented demand updater functionality in OpenPSI using MeTTa programming language.
Modified and optimized commonly used MeTTa functions, contributing to the core utilities library.
Certifications
Continuous learning through specialized courses in machine learning, computer vision, and artificial intelligence.
Coursera
Comprehensive course covering TensorFlow basics, neural networks, CNN models, and MNIST dataset training with maxpooling layers and filters.
Coursera
Advanced machine learning course covering k-means clustering, anomaly detection, reinforcement learning fundamentals, and recommender systems.
Coursera
Comprehensive computer vision course covering image processing with OpenCV, machine learning classifiers, CNNs, and object detection using Haar cascades.
Coursera
Statistical foundations course covering probability theory, distributions, hypothesis testing, Bayesian statistics, and statistical inference methods.
Coursera
Advanced NLP course focusing on RNNs, GRUs, LSTMs, bidirectional networks, named entity recognition, and Siamese networks.
Coursera
Fundamental computer science course covering searching and sorting algorithms, heap data structures, hashtables, quicksort, and quickselect algorithms.
Skills & Expertise
A comprehensive skill set spanning from traditional programming to cutting-edge AI research and neural-symbolic integration.
Proficient in multiple programming paradigms with focus on AI and systems programming.
Experience with modern web development and machine learning frameworks.
Specialized in cutting-edge AI research and cognitive system development.
Multilingual communication abilities for international collaboration.
Neural-Symbolic Integration: I'm passionate about bridging the gap between neural networks and symbolic reasoning systems. My work focuses on integrating OpenPSI cognitive architecture with neural components to create more explainable and robust AI systems.
Reinforcement Learning Integration: I believe in the potential of combining reinforcement learning with neural-symbolic systems to create AI agents that can learn, reason, and explain their decisions effectively.
Explainable AI: My goal is to develop AI systems where the symbolic component provides explainability while neural components handle pattern recognition and learning, reducing hallucinations in large language models.
Get In Touch
I'm always interested in discussing AI research, collaboration opportunities, or potential projects in neural-symbolic integration.
Whether you're interested in neural-symbolic AI research, have a project idea, or want to discuss the future of explainable AI, I'd love to hear from you.