Walter Paci

About Me

I am currently completing my PhD in Digital Humanities at the University of Florence, where I also hold a Research Fellowship. My current work contributes to an Environmental Humanities project that explores the human aesthetic relationship with nature through both qualitative and quantitative analysis. I am also a member of the research group Lablita (Laboratorio di linguistica italiana), led by Alessandro Panunzi. Alongside my academic research, I work as an AI Expert at FairyMinds.

Research Interests

  • Computational pragmatics and semantics: how Large Language Models (LLMs) detect and use pragmatic features such as irony and implicit content, and how they retrieve relevant lexical and contextual information.
  • Automatic generation and simplification of technical language: tools and models designed to improve the clarity and accessibility of professional, bureaucratic, and legal texts.
  • Cognition and philosophy of semantics: theoretical and empirical investigations into how speakers acquire, represent, and adapt meaning in context, and how this relates to cognitive and philosophical models of language use.

News

April 2026
  • I am excited to share that I have been awarded a research fellowship at the University of Florence for the coming year. I will be working with a transdisciplinary team of researchers on an Environmental Humanities project exploring the human aesthetic relationship with nature through both qualitative and quantitative analysis. I can't wait to get started! :seedling:
March 2026
  • Wonderful news! Two papers co-authored with Alessandro Panunzi and Lorenzo Gregori have been accepted at LREC 2026! 🎉 See you at the CMCL and PoliticalNLP Workshops in Palma de Mallorca :es: this May! :tropical_drink:
  • I am thrilled to announce that I will be working at FairyMinds as an AI Expert for the next three months. I am excited to apply my research and expertise in a real-world setting and contribute to the development of innovative AI solutions. Looking forward to this new adventure! :rocket:
February 2026
  • Excited to share that the RES2 team, which participated in our IMPOLS task at EVALITA 2026, received the "Federico Sangati" Best System Across Tasks Award 🏅! Congratulations to all the team members on this wonderful achievement!
October 2025
  • October will mark the final chapter of my journey as a PhD student at the University of Florence. I am thrilled to share the title of my thesis: "Linguistic Corpora and Digital Lexicographic Tools: Reasoning about Implicit Meaning in Large Language Models". I am deeply grateful to my advisors, colleagues, and friends for their support throughout this journey, and I look forward to sharing my research and celebrating this milestone with everyone! :tada:
September 2025
  • Great news! Our task IMPOLS (IMplicit contents in POLitical Speech) has been accepted at EVALITA 2026! Try the task and submit your system by the end of the year! See you in Bari next February. :it:
June 2025
  • Excited to be visiting the MaiNLP Lab in Munich :de: for the next two months to work on the automatic evaluation of implicit content explanations!
  • Grateful to be part of the organizing committee of FLUI once again for its second edition!
May 2025
February 2025
  • Valentina Saccone and I will present at the Conversazioni del GSCP workshop, where we will discuss a preliminary study on the ability of humans and LLMs to understand and explain irony in political discourse. Looking forward to seeing you in Florence! :it:
January 2025
  • Exciting news! Valentina Saccone and I had our ongoing work on irony accepted at HuMaLa. See you in Granada next June! :es:
October 2024
September 2024
June 2024
May 2024
  • Thrilled to announce that my work on LLMs' understanding of implicit content in Italian political discourse has been accepted at the second edition of AI-ROM. Can't wait to see Elba at night in Dresden! :de:
April 2024
  • It was great to give a talk on Prompt Engineering and LLMs and connect with Master's students in Digital Humanities at the University of Pisa! Thanks to Prof. Mirko Tavosanis for inviting me! :microphone:
February 2024
January 2024
  • I am excited to announce that my ongoing work investigating LLMs' ability to detect and classify implicit content has been accepted at the third edition of UnImplicit! See you at EACL 2024 in Malta. :malta:
December 2023
  • It was great to take part in the ninth edition of CliC-it. So many interesting papers to learn from!
July 2023
June 2023
May 2023
December 2022
November 2022