Lorenzo Bertolini Agnoletto
I am a research fellow at the European Commission Joint Research Centre (JRC). My current research focuses on multimodality and interpretability for biomedical NLP and machine learning, with a specific interest in graph-based representations. I hold a PhD in informatics from the Univesity of Sussex, where I was also a postdoc at the Artificial Intelligence Research Group, where my research has focused on problems at the interaction of linguistics and machine learning, such as studying and improving compositional generalisation and inference abilities of different types of NLP models and adopting language models (LLMs) to study dream content.
Relevant News
- May 2024: Our work On Constructing Biomedical Text-to-Graph Systems with Large Language Models got accepted for publication at EWCS' KNOWLEDGE GRAPH GENERATION FROM TEXT (TEXT2KG) workshop.
- April 2024: Invited talk at the CBS Research Seminar of The Hong Kong Polytechnic University, titled "Field of dreams: applying NLP to automatize dream content analysis"
- Junary 2024: Our work Automatic Annotation of Dream Report’s Emotional Content with Large Language Models got accepted for publication at EACL 2024 Computational Linguistics and Clinical Psychology workshop.
- September 2023: I joined the European Commission Joint Research Center (JRC) as a research fellow in NLP.
- May 2023: DReAMy has been updated to version 1.5 🥳! Check out the Git repo for Colab/notebook tutorials, and the Twitter post for a quick summary of the new features.
- April 2023: Our work Towards Unsupervised Compositional Entailment with Multi-Graph Embedding Models was accepted for publication at the 15th IWCS conference as a long paper.
- April 2023: Our extended abstract for the DReAMy toolkit was accepted for presentation at the World Sleep Congress 2023.
- 16 Feb 2023: I defended my PhD Thesis, and passed with minor corrections 🥳.
- August 2022: Our work Testing Large Language Models on Compositionality and Inference with Phrase-Level Adjective-Noun Entailment was accepted for publication at COLING 2022 as a long paper.
- April 2022: Bocconi's MilaNLP was beyond nice and invited me to rant about compositionality in language models in their weekly lab seminar.
- October 2021: I gave an invited talk at the University of Sussex's Data Science Society on my ongoing PhD work and the phenomena of Compositionality in humans and machines.
- May 2021: Our work Representing Syntax and Composition with Geometric Transformations was accepted for publication at Findings of ACL 2021 as a long paper.