about
Hi 👋 I'm Vincenzo di CiccoMy interests range from pure Software Development to Web Security, from Machine Learning to Data Science. CurrentlyI'm in Rome, working on Large Language Models for Translation.
posts
-
no posts yet ¯\_(ツ)_/¯
research and activities
- Ranked 3rd in the comma.ai CTF
- Most Innovative award in the Twitter's Machine Learning Algorithmic Bias Bounty Challenge [code]
- CVE-2021-22540 - XSS vulnerability in dart:html
- CVE-2020-8923 - XSS vulnerability in dart:html
- Interpreting deep learning models for entity resolution: an experience report using LIME. Vincenzo Di Cicco, Donatella Firmani, Nick Koudas, Paolo Merialdo, Divesh Srivastava. In Proceedings of aiDM@SIGMOD 2019: 8:1-8:4
projects and contributions
- EduFSDP - a minimal, educational implementation of Fully Sharded Data Parallel (FSDP).
- Tiny gradcheck @geohot/tinygrad
- Carbonara Extractor - Product specifications extractor driven by a DOM Classifier.
- Mojito - Interpreting deep learning models for entity resolution.
- Cwatch - A lightweight filesystem monitor able to trigger user-defined commands.
- Mob - A PoC of a decentralized and fault-tolerant network for the deployment and scaling of services based on Kademlia.
- Kademlia - an old and incomplete implementation of the Kademlia DHT in Erlang.