Skip to content
Atlas profile

Pop Alexandru

@pop123-uxVerified GitHub profile

Founder, Student & Aspiring AI Researcher | ICYS 2026 Gold Medalist | Building AI projects, researching independently and starting my career

AI-Atlas score

Specialist
0100
Strongest
Machine Learning
73.1/100 domain score
Rank
4th=
of 86
Romania
1st
of 9
  • RomaniaLocation
  • 12 Nov 2026Mapped
  • 30 / 41AI projects
01Skills matrix

AI Skills Matrix

AI Skills Matrix scores, 0 to 100
Machine Learning73
Computer Vision69
Data Science42
NLP & LLMs67
AI Agents33
MLOps14

Each node is one domain, scored only from evidence in these public repositories. Select one to see why.

Deep, repeated, well-engineered work in this domain.

02Evaluated projects6 of 30

41 public repositories → 30 with AI work, evaluated. The other 11 (repositories without AI code, forks and empty repositories) are not scored and have no effect on the result. How repositories are selected

Ranked by contribution

Ordered by how much each project adds to the Atlas score, not by its own score. A project with evidence in several domains can rank above one with a slightly higher project score. Choose a domain to focus on the projects with meaningful evidence there.

01
67.4

Advanced

20% of Atlas Score

Library / Framework· Python

ORBIT's cover

The ORBIT repository provides a Python implementation of the ORBIT optimizer for transformer models with rotary positional embeddings. It includes the optimizer, adapters for external models, baselines, training scripts, and detailed documentation for reproduction and use.

Machine Learning0.90NLP & LLMs0.80
arxiv-papersdeep-learningmuonoptimizerresearch
02
66.0

Advanced

13% of Atlas Score

Training Experiment· Python

ORBIT-all is a research repository implementing a function-aware optimizer for language model training, complete with a custom transformer-like model, experiment harness, and detailed protocol for ablation studies and validation.

Machine Learning0.90NLP & LLMs0.80
03
60.5

Advanced

11% of Atlas Score

Architecture Reimplementation· Jupyter Notebook

image

The repository provides a from-scratch PyTorch implementation of the AlexNet convolutional neural network, including model definition, training loop, evaluation, and a pretrained checkpoint on the Imagenette dataset. It is intended as an educational reference and demonstration of reproducing a classic CNN architecture.

Computer Vision0.90Machine Learning0.80Data Science0.20MLOps0.20
alexnetalexnet-architecturealexnet-cnnalexnet-modelalexnet-models
04
58.6

Advanced

8% of Atlas Score

ML Application· Python

This repository provides a PyTorch-based implementation of neural style transfer using a pretrained VGG19 model. It includes a CLI for transferring the style of one image onto another, along with unit tests and documentation.

Machine Learning0.90Computer Vision0.90MLOps0.20
cnncnn-architecturecnn-classificationcnn-modelcnn-pytorch
05
62.2

Advanced

8% of Atlas Score

Architecture Reimplementation· Jupyter Notebook

ResNet Family project header

This repository provides a from-scratch reimplementation of the ResNet family (ResNet-12, ResNet-18, ResNet-34, ResNet-50) in PyTorch, with manual implementations of BatchNorm and activation functions for educational purposes, including training and evaluation on CIFAR100.

Machine Learning0.90Computer Vision0.80Data Science0.30MLOps0.20
aicifar100cifar100-image-classificationcnncomputer-vision
06
57.7

Implementation

6% of Atlas Score

Architecture Reimplementation· Jupyter Notebook

LeNet-5 project header

This repository provides a from-scratch PyTorch reimplementation of the LeNet-5 architecture, including a custom radial basis function output layer and loss function as described in the original paper. It includes unit tests, a training notebook, and a trained checkpoint for MNIST digit classification.

Machine Learning0.90Computer Vision0.90Data Science0.20MLOps0.20
aiartificial-intelligencedeep-learningdeep-neural-networksdeeplearning