.png)
In the first two years of the project, the project consortium partners already published 14 papers mainly, but not only, focusing on large language models (LLM) and other technologies used to produce and upgrade tools within the VIGILANT Platform.
9 of the papers have been authored or co-authored by the Kempelen Institute of Intelligent Technologies (KInIT), 4 by the University of Sheffield, one by OntoText and one by the University of Freiburg. Find your favourite in the list!
Papers produced by KInIT:
Fighting Randomness with Randomness: A novel "DENI" strategy
IMGTB: A Framework for Machine-Generated Text Detection Benchmarking
Authorship Obfuscation in Multilingual Machine-Generated Text Detection
Disinformation Capabilities of Large Language Models
Fine-tuned LLMs for Multilingual Machine-Generated Text Detection
Enhancing Multilingual Fine-Tuning for the Detection of Persuasion Techniques
MULTITuDE: A New Benchmark for Multilingual Machine-Generated Text Detection
Papers produced by the University of Sheffield:
Analysing State-Backed Propaganda Websites: A New Dataset and Linguistic Study
Mono and Multilingual Approaches for News Genre, Topic and Persuasion Technique Classification
Comparative Study on Accurate COVID-19 Information vs. Misinformation
Paper co-authored by KInIT and the University of Sheffield:
Comparison Between Parameter-Efficient Techniques and Full Fine-Tuning
Paper by Dr. Elisa Orru, University of Freiburg:
Are Publicly Available (Personal) Data 'up for grabs”? A Discussion of Three Privacy Agreements
Paper by Ontotext:
NEXT: An Event Schema Extension Approach for Closed-Domain Event Extraction Models