My beautiful image

steve.azzolin {at} unitn {dot} it

Ciao 👋

My name is Steve Azzolin and I'm currently an ELLIS PhD student co-supervised by Andrea Passerini (University of Trento), Bruno Lepri (FBK), and Pietro Liò (University of Cambridge). Starting from 2024, I'm also an organiser of the LoG Conference.

My PhD is focusing on studying how to make machine learning algorithms more reliable.

INTERESTS: Graph Neural Networks; Representation Learning; Explainability

EDUCATION: BSc in Computer Science; MSc in Artificial Intelligence Systems


Spotlights

AIT Journal Club

Check the website to stay up-to-date with the last events

News

Let's stay in touch on X

Joined the organisation team of LoG Conference
Curr.
Organizing the Trento LoG meetup
11/27-11/30 2023
Hand-on tutorial on XAI for GNNS @ TuringInstitute
11/16/2023
Starting of my PhD
11/01/2023

Publications

For an up-to-date list, check out my Scholar

Sheaf Diffusion Goes Nonlinear: Enhancing GNNs with Adaptive Sheaf Laplacians
O. Zaghen, A. Longa, S. Azzolin, L. Telyatnikov, P. Liò, A. Passerini
[GRaM Workshop @ ICML 2024]
2024
Explaining the Explainers in Graph Neural Networks: a Comparative Study
A. Longa, S. Azzolin, G. Santin, G. Cencetti, P. Liò, B. Lepri, A. Passerini
[ACM Computing Surveys]
2024
A Simple Latent Variable Model for Graph Learning and Inference
M. Jaeger, A. Longa, S. Azzolin, O. Schulte, A. Passerini
[LOG2023]
2023
Global Explainability of GNNs via Logic Combination of Learned Concepts
S. Azzolin, A. Longa, P. Barbiero, P. Liò, A. Passerini
[ICLR2023]
2023
Multi-source Multi-domain Sentiment Analysis with BERT-based Models
G. Roccaruna, S. Azzolin, G. Riccardi
[LREC 2022]
2022
Can Emotion Carriers Explain Automatic Sentiment Prediction? A Study on Personal Narratives
S.M. Mousavi, G. Roccabruna, A. Tammewar, S. Azzolin, G. Riccardi
[WASSA ACL Workshop]
2022

Talks

XAI4GNNs Hands-on Tutorial

Hands-on Tutorial of XAI4GNNs held at the Turing Institute.

Global Explainability of GNNs via Logic Combination of Learned Concepts

Shorter version of my prev. Cambridge Talk held at the AI Trento Journal Club.

Global Explainability of GNNs via Logic Combination of Learned Concepts

Presented our latest work on Global Explainability of GNNs via Logic Combination of Learned Concepts at the University of Cambridge as part of the Artificial Intelligence Research Group Talks series.