Language Breakdown
Lines of code distribution across 19 owned repositories
I-Shaped Developer
I-shapedSpecialist — deep expertise in Jupyter Notebook
Collaboration Network
Global Impact visualization
Repos
24
PRs
0
Growth
+18%
Top Collaborators
No collaborator data yet.
Coding Streak
Contribution activity over the past year
Jimmy Lin
@lintool
Ali Moradzade
@ali-moradzade
Pouya Taghipour
@Pouya-Ta
Ali Noroozbeigy
@Ali-Noroozbeigy
Mahdieh Sadatbenis
@smahdieh
Top Repositories
Fine-tuning the xlm-roberta-base model on the PQuAD dataset, designed for question answering in the Persian language. The project builds on Hugging Face's original notebook, adapting it to work with a Persian dataset and evaluating the model's performance with metrics like Exact Match and F1 Score.
A collection of Python scripts demonstrating the use of Captum for visualizing and understanding deep learning model predictions on the CIFAR-10 dataset. Includes implementations of Grad-CAM, Integrated Gradients, and Gradient SHAP for interpretability.
A sophisticated information retrieval system focused on news articles, featuring advanced document indexing, querying, and relevance ranking mechanisms. Implements TF-IDF weighting and vector space models to enhance retrieval precision and recall, with optimized search algorithms for improved user satisfaction.
A simulation project utilizing fuzzy logic to mimic decision-making processes in self-driving cars. The simulator is designed to manage various driving conditions, with a focus on optimizing fuzzy rules for obstacle avoidance and speed control.
My personal website, which showcases my CV
This repository evaluates the impact of SHAP, DeepLIFT, and GradientSHAP-based regularization on the robustness of neural networks trained on the MNIST dataset.
This repository contains a Python script that uses the Captum library to perform an initial analysis of various attribution methods on a ResNet18 model trained on the CIFAR-10 dataset. It evaluates and compares the performance of different interpretability techniques through infidelity and sensitivity metrics.
Analyze the impact of background regions on image classification using Gradient SHAP. This project visualizes how a ResNet18 model, fine-tuned on CIFAR-10, leverages different parts of an image for decision-making, providing insights into model interpretability.
Open Source Impact
Contributions to external projects
No external contributions found.