RAG Chatbot
Used OpenAI's ChatGPT-3.5 with Retrieval-Augmented Generation (RAG) on a collection of recent and personally selected AI safety papers.
Jul 20, 2025
Jail breaking LLMs
A small project were I redteam differetn LLMs models with type of attacks.
May 26, 2025
Out-of-Distribution (OOD) Detection with Adversarial Probes
Leveraging adversarial perturbations as diagnostic probes to enhance the performance of models on out-of-distribution (OOD) detection.
May 6, 2025
3D Adversarial Camouflage
A high-level overview of 3D physical adversarial attacks designed to fool AI object detection systems in real-world conditions.
Dec 21, 2024
Adversarial Attacks for Facial Verification Systems
A high-level overview of black-box adversarial attacks against facial verification systems under realistic constraints.
Oct 26, 2023
Active Learning
Different Active Learning strategies for selecting the most informative images to train object dectors models.
Oct 26, 2023