Francesca Padovani

News

• August-September 2026
  Last Translation Benchmark v1 is out!
  I enjoyed so much being involved into the first release of the Last Translation Benchmark both as a contributor and as a reviewer. You can read the manuscript at Arxiv.
• August 2026
  Non-archival paper accepted at BabyLM Workshop 2026!
  We introduce an ongoing work I have been involved with Bastian Bunzeck on comparing synthetic (automatically generated) vs natural child-caregiver conversations. We peopose a new instruction-tuning generation method with data created from real dialogue.
• May 2026
  New pre-print out. Paper accepted at CLiC-it 2026 (oral presentation)
  Together with Sara Candussio and Daniel Scalena we had fun evaluating LLMs on the Taboo game in Italian :) Check out it on Arxiv.
• April 2026
  New pre-print out!!
  Together with Malvina Nissim (GroNLP) we are starting structurally evaluating the experience of assembling a puzzle about conversational chatbots and GenAI and in the meantime we are introducing the project with an auxiliary paper. Have a look at the website.

Hi! I am a second year PhD Student of the Computational Linguistic Group at the Center for Language and Cognition (CLCG), the University of Groningen (NL). My research is conducted under the supervision of Arianna Bisazza and the co-supervision of Yevgen Matusevych and Jaap Jumelet. The VIDI project I am working on, funded by the Dutch Research Council (NWO), focuses on improving language modeling for (low-resource) morphologically rich languages, taking inspiration from child language acquisition insights.

My research experience has been quite diverse, in line with the hybrid nature of my education. I have a humanistic background, since did classical studies during my high school (learning Latin and Ancient Greek). I then completed my Bachelor's degree in Psychology at the University of Trieste (Italy) and carried out my undergraduate internship in the Language,Learning and Reading Lab of the Cognitive Neuroscience Unit at SISSA, led by Davide Crepaldi. I then pursued a Master's degree in Data Science at the University of Trento (Italy). For my Master thesis I spent 6 months in the Language group of the Computer Science Lab Sony in Paris, where I was supervised by Martina Galletti and Remi van Trijp as part of a project aimed at building a multimedia interface to support Italian children with Learning Disabilities during therapy with speech therapists.