Michele Papucci

Italian Multilevel Parallel Corpus for Text Simlification, or IMPaCTS for short. 1.066.828 fully generated simplification for Wikipedia and Public Administration texts automatically annotated for readability levels. For each human-written original sentence, up to 10 different simplifications are provided.

Repository for the LLMs Anatomy Course held by me and Alessio Miaschi for the University of Pisa Engineering Department

Uncertainty Gated Lexical Decoding (UGLD) library repository.

TextGenerator is a PHP package that aims to generate automated texts from data.

Progetto di codifica di testi 2020/2021 di Michele Papucci e Sofia Zuffi

Personal Website hosted on Github Pages

No description.

No description.

No description.

Project for the creation of a biologic medicinal transportation device.

The repository contains the dataset created for an in-depth linguistic analysis of automatically generated text simplification for Italian using Llama-2.

Probabilistic Neural Network's Exam 2025

A Quartz fork, with a new component to show any kind of Frontmatter Obsidian Property on the pages.

Experiments on using attention mechanisms for label selection in T2T model for classification

Scraper for the vernacoliere. A satiric newspaper from Livorno, Italy.

PyTorch implementation of char-rnn (character-level language model)

Exam's project for distributed data analysis and mining exams

Data Mining project for the 2021/2022 unipi course of Data Science and Business Informatics.

Double pendulum simulation in js