ART: Artificial Intelligence Research at Tor Vergata

ART: Artificial Intelligence Research at Tor Vergata's repositories

KERMIT

🐸 KERMIT - A lightweight library to encode and interpret Universal Syntactic Embeddings

Language:JavaScriptMIT57 6 1

GASP

GASP! Dataset - Generating Abstracts of Scientific Papers from Abstracts of Cited Papers

NOASSERTION9 60

C-OSINT

Language:Jupyter Notebook1 20

HateSpeechKermit

KERM-HATE push forward the research on how syntactic information can be used to de-bias hate speech recognizers and so to contribute to solve problems of prejudice.

Language:Jupyter NotebookMIT000

Ita_CurriculumLearning

This repository contains the code of the paper: Are All Languages Equal? Curriculum Learning over Different Languages

Language:Jupyter Notebook020

KERMIT-4-Sentiment-Analysis-on-Italian-Reviews-in-Healthcare

Language:Jupyter Notebook030

KERMIT4NLI

Language:Jupyter Notebook000

P-AT

Data and code for "Measuring bias in Instruction-Following models with P-AT"

Language:PythonGPL-3.0020

PreCog

This repository contains the code of the paper: PreCog: Exploring the Relation between Memorization and Performance in Pre-trained Language Models

020

Smart-visualization-of-Heat-Parse-Trees-for-KERMIT

Creation of a function that takes in input an active tree and draws his smart visualitation into a canvas

Language:HTML000

Transformers_DarkWeb

This repository contains the code of the paper: The Dark Side of the Language: Pre-trained Transformers in the DarkNet

000