Davide Ghilardi
MSc student at University of Milano Bicocca.
I’m a MSc Data Science student and fellow NLP researcher at the Insid&s Lab at the University of Milano Bicocca under the supervision of Matteo Palmonari, working on saefty and interpretability of LLMs.
I’ve been a VSR at Stanford under the supervision of Dan Jurafsky and Federico Bianchi working on mechanistic interpretability for AI Safety. Recently, I’ve been working on validating and improving Sparse Autoencoders.
I’m also a volunteer ML specialist at The Good Scientists.
news
| Nov 15, 2024 | I’m presenting our new work “Accelerating Sparse Autoencoder Training via Layer-Wise Transfer Learning in Large Language Models” at BlackBoxNLP at EMNLP 24’ in Miami. |
|---|---|
| May 01, 2024 | On behalf of TheGoodScientists, I’m at META HQ in Menlo Park as a representative of the Llama Impact Community for Social Impact. |