Kavit Tolia
The speaker spent over 12 years working in quantitative roles in investment management before returning to academia to study Artificial Intelligence. They are currently completing a Master’s degree in AI and ML in Science, and are particularly interested in how modern machine learning systems behave in practice, especially where modelling assumptions quietly break down.
Session
06-06
15:30
45min
Do Multilingual Embeddings Really Share a Semantic Space? Practical Lessons Across Scripts and Languages
Kavit Tolia
Multilingual language models are often assumed to embed different languages into a shared semantic space. This talk presents empirical results showing how script, tokenisation, and data imbalance shape multilingual embeddings in practice, and offers practical diagnostics for evaluating their reliability before deployment.
Hardwick Hub