Research & Projects
📚Selected Publications
Now You See Me (CME): Concept-based Model Extraction
D Kazhdan, B Dimanov, M Jamnik, P Lio, A Weller
MEME: Generating RNN Model Explanations via Model Extraction
D Kazhdan, B Dimanov, M Jamnik, P Lio
MARLeME: A Multi-Agent Reinforcement Learning Model Extraction Library
D Kazhdan, Z Shams, P Lio
GCI: A (G)raph (C)oncept (I)nterpretation Framework
D Kazhdan, B Dimanov, LC Magister, P Barbiero, M Jamnik, P Lio
Manipulating SGD with Data Ordering Attacks
I Shumailov, Z Shumaylov, D Kazhdan, Y Zhao, N Papernot, MA Erdogdu, R Anderson
Algorithmic Concept-based Explainable Reasoning
D Georgiev, P Barbiero, D Kazhdan, P Velickovic, P Lio
🚀Selected Research & Engineering Projects
Concept-based XAI (CXAI) framework
A comprehensive framework for concept-based explainable AI approaches, implementing several SOTA methods for Concept-based XAI and enabling their comparisons + improvements.
CME: a Concept-based Model Extraction Framework
A framework for extracting interpretable concept-based models from black-box neural networks, enabling better understanding of AI decisions.
MARLeME: a Multi-Agent Reinforcement Learning Model Extraction Library
A multi-agent reinforcement learning model extraction library for understanding complex AI systems in multi-agent environments.
MEME: a Model Explanation via Model Extraction Framework for RNN Explainability
A Model Extraction framework for RNN explainability, providing interpretable explanations for recurrent neural network decisions.
GCI: a Graph Concept Interpretation Framework
A framework for graph concept interpretations, enabling explainable AI for graph neural networks and graph-based machine learning.