Computational tools
  1. 2026 - Small Language Models (v2026)
    Five trained SLMs based on a 3M token corpus in Italian and on the 10M BabyLm corpus in English
     
  2. 2025 - MorPiece (v1.3.1)
    An updated split-based tokenization library that incrementally segments words into potentially meaningful morphemes
     
  3. 2024 - BabyLM models (2024)
    preprocessing, tokenization, models architectures for Small Language Modeling
     
  4. 2023 - Expectation-based Minimalist Grammars (v1.0)
    An algorithmic implementation of a Minimalist Top-Down derivation with on-line complexity predictions
     
  5. 2023 - BEXT tool (v0.2)
    A tool for Extracting raw PPG data from wearable sensors (in collaboration with Mauro Dragoni, FBK)