captum

Interpretability for sequence generation models 🐛 🔍

We introduce XBrainLab, an open-source user-friendly software, for accelerated interpretation of neural patterns from EEG data based on cutting-edge computational approach.

End-to-end toxic Russian comment classification

This repository contains the source code for Indoor Scene Detector, a full stack deep learning computer vision application.

Deep Classiflie is a framework for developing ML models that bolster fact-checking efficiency. As a POC, the initial alpha release of Deep Classiflie generates/analyzes a model that continuously classifies a single individual's statements (Donald Trump) using a single ground truth labeling source (The Washington Post). For statements the model d…

🔍 Enhance medical imaging with a lightweight CNN model that offers over 91% accuracy and integrated explainability for better clinical trust.

OdoriFy is an open-source tool with multiple prediction engines. This is the source code of the webserver.

🔬 Deep-Viz: Unveiling the Black Box of Deep Learning

A lightweight Explainable AI CNN for PathMNIST medical imaging, achieving 91%+ accuracy with Integrated Gradients and SQLite-based attribution storage. Built in PyTorch, this scalable model delivers high performance, transparency, and real-world readiness, making it ideal for medical AI, edge deployment, and explainable deep learning research.

An AI capable of running 100% locally using a hybrid TCN-LSTM-Attention model to predict time sequences in a way that is understandable to everyone through open-source LLMs

🚗 Analyze and visualize decision-making in autonomous driving RL agents using Integrated Gradients for clearer interpretability in complex driving tasks.

VisionDriveX is a multi-task autonomous driving perception system that performs traffic sign classification, stop-sign detection, and lane segmentation. Built with PyTorch and explainable AI (Grad-CAM), it delivers real-time, interpretable road understanding for safety-critical ADAS applications.

Breast cancer metastasis detection in lymph node histopathology images using ResNet-18, slide-independent 5-fold cross-validation, Grad-CAM, and Integrated Gradients.

Interpretability in Autonomous Driving: Visual Attribution Analysis of RL Agents

An AI-driven system for detecting dark patterns in user interfaces, enhanced with explainability to clearly interpret why a pattern is flagged. Designed to improve transparency, trust, and ethical design in digital products.

Explaining a spiking neural network on MNIST with feature attribution, including how its decision builds up over time.

Interpretability Metrics

Whether a vision model explanation survives an attack on the image. Course at the University of Trieste.

Explainable AI benchmark (IG, WINIT, LIME, PI) for CNC machine energy prediction.

Cat vs dog computer vision experiment with PyTorch, MobileNetV3, MLflow tracking, data augmentation, error analysis and explainable AI