Connectionists: Live free 'AIDA/HiTZ AI Excellence e-Lecture' by Associate Prof. Ekaterina Shutova: "Cross-lingual information sharing in multilingual language models", October 24, 2024 - 15:00 CET

[Efi]

Dear AI scientist/engineer/student/enthusiast,



*Associate Professor Ekaterina Shutova (*University of Amsterdam) will
deliver the e – lecture: “*Cross-lingual information sharing in
multilingual language models*” on *24th October 2024, 15:00 CET.*



Multilingual language models (MLMs), such as XLM-R or BLOOM, are pretrained
on data covering many languages and share their parameters across all
languages. This modeling approach has several powerful advantages, such as
allowing similar languages to exert positive influence on each other, and
enabling cross-lingual task transfer (i.e., fine-tuning on some source
language(s), then using the model on different target languages). The
success of such transfer, however, depends on the model's ability to
effectively share information between different languages in its parameter
space. Yet, the cross-lingual information sharing mechanisms within MLMs
are still not fully understood. In this talk, I will present our recent
research that investigates this question from three different perspectives:
encoding of typological relationships between languages within MLMs,
language-wise modularity of MLMs and the influence of training examples in
specific languages on predictions made in others.

   - For more details, click *here
   <https://www.i-aida.org/events/cross-lingual-information-sharing-in-multilingual-language-models/>*
   - Registration [*here <https://www.hitz.eus/webinar_izenematea>*].





The *International AI Doctoral Academy (AIDA) <http://www.i-aida.org/>*, a
joint initiative of the European R&D projects *AI4Media
<https://ai4media.eu/>*, *ELISE <https://www.elise-ai.eu/>*, *Humane AI Net
<https://www.humane-ai.eu/>*, *TAILOR <https://tailor-network.eu/>, VISION
<https://www.vision4ai.eu/>*, is very pleased to offer you top quality
scientific lectures

in the framework of *AIDA AI Excellence Lecture Series
<https://www.i-aida.org/event_cat/ai-lectures/>* on several current
hot AI topics. Lectures will be offered alternatingly by:

   - Top highly cited senior AI scientists internationally or
   - Young AI scientists with promise of excellence (AI sprint lectures)

These lectures are disseminated through multiple channels and email lists
(we apologize if you received it through various channels). If you want to
stay informed on future lectures, you can register in the email lists *AIDA
email list <https://lists.auth.gr/sympa/info/aida>* and *CVML email list
<https://lists.auth.gr/sympa/info/cvml>*.



Upcoming AI Excellence Lectures:

   - Dr. Hannes Fassold, Joanneum Research, “Manifold / geometric deep
   learning and its application for multimedia”, 12/11/24 – 17:00 CET
   - Prof. Dr. Gerard Pons-Moll, “How to train large scale 3D human
   foundation models”, 19/11/24 – 17:00 CET
   - Associate Prof. Gül Varol, “Text-driven 3D human motion synthesis”,
   26/11/24 – 17:00 CET
   - Prof. Sean Gong, “LLMs for More Reliable Small Data Learning”, 3/12/24
   – 17:00 CET



Best regards
Profs. N. Sebe, M. Chetouani, P. Flach, B. O’Sullivan, I. Pitas, J.
Stefanowski
AIDA AI Excellence Lecture Series committee members
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.srv.cs.cmu.edu/pipermail/connectionists/attachments/20241018/226df668/attachment-0001.html>