Matching Pairs: Attributing Fine-Tuned Models to their Pre-Trained Large Language Models
- Myles Foley
- Ambrish Rawat
- et al.
- 2023
- ACL 2023
Bio
Giulio's research interests focus on the security and robustness of machine learning. Recently, he has been working on adversarial machine learning with a focus verifiable methods and the security of Federated Learning.
He obtained his PhD from Imperial College London in 2021 which focuses on adversarial and robust machine learning. He joined IBM Research in 2021 and has continued his research in private and robust ML and contributing to open source projects such as ART (Adversarial Robustness Toolbox) and various Horizon Europe programmes.
Publications
Machine Learning Platform for Extreme Scale Computing on Compressed IoT Data
- 2022
- Big Data 2022
Federated Continual Learning with Differentially Private Data Sharing
- 2022
- NeurIPS 2022
Verifiable Federated Learning
- Simone Bottoni
- Giulio Zizzo
- et al.
- 2022
- NeurIPS 2022
Robust Learning Protocol for Federated Tumor Segmentation Challenge
- 2022
- MICCAI 2022
FAT: Federated Adversarial Training
- 2020
- NeurIPS 2020
Projects
Incremental Machine Learning for Extreme Scale Computing
AutoML for incremental machine learning algorithms for big time-series data.
Top collaborators
AR
Ambrish Rawat
Senior Research Scientist, AI Security and Privacy, Master Inventor
ST
SB
Stefano Braghin
Research Engineer and Technical Lead - Security and Privacy, Master Inventor
DS
Dhaval Salwala
Research Engineer | ML and Extreme Scale Computing