Adaptive Natural Language Processing Using Large Language Models - PoVeJMo
Principal Investigator at ZRC SAZU
Janoš Ježovnik, PhD-
Original Title
Prilagodljiva obdelava naravnega jezika s pomočjo velikih jezikovnih modelov – PoVeJMo
Project Team
Dr. Duša Divjak Race, Dr. Dejan Gabrovšek, Dr. Januška Gostenčnik, dr. Mateja Jemec Tomazin, Dr. Janoš Ježovnik, Dr. Karmen Kenda-Jež, Dr. Domen Krvina, Dr. Mojca Kumin Horvat, Dr. Nina Ledinek, Dr. Mija Michelizza, Dr. Tanja Mirtič, Dr. Andrej Perdih, Dr. Špela Petric-Žižić, Dr. Jožica Škofic, Ema Štarkl, Dr. Mitja Trojar-
Duration
1 September 2023–30 June 2026 -
Lead Partner
-
Project Leader
-
Financial Source
NextGenerationEU
Načrt za okrevanje in odpornost
ARIS - Javna agencija za raziskovalno dejavnost RS
Partners
CJVT - Center za jezikovne vire in tehnologijo, FRI - Fakulteta za računalništvo in informatiko, Inštitut za novejšo zgodovino, ZRC SAZU
The key objective of the long-term research and innovation (R&I) program Adaptive Natural Language Processing Using Large Language Models (PoVeJMo) is the development of large language models that influence nearly the entire field of artificial intelligence and machine learning, while also having a significant impact on many other areas and on society as a whole. The new open-access and computationally efficient language models will serve as the foundation for advanced applications in medicine, the humanities, industrial environments, and software development. Large generative language models and their adaptation for instruction-following and dialog-based communication will also provide the core infrastructure for artificial intelligence applications in the Slovenian language.