When? Thursdays at 16:30 (on specific dates from September to November).
Where? In the main building of the University of Latvia (Raiņa Bulvāris 19), auditorium 16.
The next guest lecture is expected this Thursday, October 2 at 16:30, when the audience will be introduced to practical solutions of transformer architecture: from BERT to LLM by Dmitrijs Jemeļjanovs, who is the head of LLC “Forward IT Consulting” and a member of the Council of the Latvian Association of Artificial Intelligence.
List of all guest lectures:
25.09. Jurgita Kapočiūtė-Dzikienė, PhD. “Large Language Models: Their Internal Architectures and Differences”
- Professor at the Faculty of Informatics of Vytautas Magnus University and Tilde Lithuania’s senior and leading researcher in language technologies
02.10. Dmitrijs Jemeļjanovs. “Transformer Architecture in Practical Solutions: From BERT to LLM”
- Head of LLC “Forward IT Consulting” and a Member of the Council of the Latvian Association of Artificial Intelligence.
16.10. Matīss Rikters, PhD. “Machine Translation, Attention Mechanism and Synthetic Data”
- Researcher at the Artificial Intelligence Research Center of the National Institute of Advanced Industrial Science and Technology (Japan)
23.10. Matīss Rikters, PhD. “Multimodal Language Models and Their Use”
- Researcher at the Artificial Intelligence Research Center of the National Institute of Advanced Industrial Science and Technology (Japan)
20.11. Toms Bergmanis, PhD. “TildeOpen LLM: A European Multilingual Model”
- Researcher at "Tilde"
27.11. Tanel Alumäe, PhD. “Speech Recognition” (in ENGLISH)
- Tenured associate professor of TalTech (Estonia)
The first guest lecture was successfully held on September 25, when Dr.sc.comp. Jurgita Kapočiūtė-Dzikienė, - professor at the Faculty of Informatics of Vytautas Magnus University and Tilde Lithuania's senior and leading researcher in language technology - gave a lecture on large language models (LLM). The lecture examined the internal mechanisms of LLMs, such as refinement approaches, chain reasoning, retrieval-augmented generation (RAG), which allow them to reason and generate better answers. It was also discussed how different LLM families differ and what these differences mean in practice.


Guest lectures can be attended by Computer Science students at all study levels, but also other interested parties!

Guest lectures are implemented with the financial support of VTI. The project “Language Technology Initiative” (No. 2.3.1.1.i.0/1/22/I/CFLA/002) is co-financed by the European Union Recovery and Resilience Mechanism Investment and the State Budget.