Skip to main content
Stefano Civelli beside a kangaroo

PhD researcher in artificial intelligence at UQ

I am a PhD researcher in Computer Science at The University of Queensland (opens in a new tab). My work examines how large language models represent difficulty and how their behaviour is shaped by context, language, and political bias.

Before starting my PhD, I worked as a Research Assistant at UQ and as an ML engineer at ML cube (opens in a new tab) in Milan. I hold an M.Sc. in Computer Science and Engineering from Politecnico di Milano (opens in a new tab).

I also teach computer science at UQ.

Selected publications

Recent Research

A Shared Geometry of Difficulty in Multilingual Language Models

Stefano Civelli, Pietro Bernardelle, Nicolò Brunello, Gianluca Demartini

ACL 2026

arXiv (opens in a new tab)

Ideology-Based LLMs for Content Moderation

Stefano Civelli, Pietro Bernardelle, Nardiena A Pratama, Gianluca Demartini

TIST - Special Issue on Risks and Unintended Harms of Generative AI Systems · 2026

arXiv (opens in a new tab)

Political Advertising on Facebook During the 2022 Australian Federal Election: A Social Identity Perspective

Stefano Civelli, Pietro Bernardelle, Frank Mols, Gianluca Demartini

ICWSM 2026

arXiv (opens in a new tab)

Experience and education

Resume

  1. PhD in Artificial Intelligence at The University of Queensland (opens in a new tab)

    Researching core aspects of Large Language Models (LLMs) like prompt complexity and bias. Developing novel methods to measure and predict query complexity for LLMs.

  2. Research Assistant at The University of Queensland (opens in a new tab)

    Conducted research on LLMs for classification of harmful content. Implemented multimodal ML models in PyTorch. Analyzed Facebook ads for political campaigns. Deployed AWS-based dashboard for campaign analysis.

  3. ML Engineer Intern at ML cube (opens in a new tab)

    Developed RL solution for AGV mission time estimation. Implemented classic and distributional RL algorithms for time estimation.

  4. M.Sc. in Computer Science and Engineering at POLIMI (opens in a new tab)

    Graduated with 110L/110. Main courses: Machine Learning, Neural Networks, Distributed Systems, Data Streaming, Recommender Systems.

    DistinctionMerit-based scholarship · Two consecutive years

  5. B.Sc. in Computer Science and Engineering at POLIMI (opens in a new tab)

    Graduated with 109/110. Main courses: software engineering, databases, algorithms & data structures, statistics, linear algebra.

    DistinctionBest Freshman Award · 2018–19