Go directly to page contentent
TEAM LADE
Lorenzo Basile
Researcher at Laboratory of Data Engineering

I am a postdoctoral researcher at Area Science Park in Trieste, Italy, having recently completed my PhD at the University of Trieste. My research focuses on interpretability and representation learning in neural networks, with a particular interest in large multimodal and vision-language models. I study how these models organize and process information internally, which components specialize across modalities, and how we can leverage these insights to make models more trustworthy, transparent, and efficient.

Research Interests
  • AI Interpretability
  • Multimodal Learning
  • Representation Learning
Experience & Education
  • PhD in Artificial Intelligence, University of Trieste, 2025
  • MSc in Data Science, University of Trieste, 2021
  • BSc in Computer Engineering, University of Naples Federico II, 2019
Latest pubblications
02/06/2026
Visual Instruction Tuning Aligns Modalities through Abstraction
Abstract Visual instruction tuning effectively adapts a pre-trained Large Language Model (LLM) to process image…
Go to the news Visual Instruction Tuning Aligns Modalities through Abstraction
18/09/2025
Head Pursuit: Probing Attention Specialization in Multimodal Transformers
Abstract Language and vision-language models have shown impressive performance across a wide range of tasks,…
Go to the news Head Pursuit: Probing Attention Specialization in Multimodal Transformers
12/04/2025
ResiDual Transformer Alignment with Spectral Decomposition
Abstract When examined through the lens of their residual streams, a puzzling property emerges in…
Go to the news ResiDual Transformer Alignment with Spectral Decomposition
18/09/2025
The Narrow Gate: Localized Image-Text Communication in Native Multimodal Models
Abstract Recent advances in multimodal training have significantly improved the integration of image understanding and…
Go to the news The Narrow Gate: Localized Image-Text Communication in Native Multimodal Models