/
Beyond Token-based LLMs: Reasoning over Multilingual and Multimodal Embeddings
Data is an LLM's Source Code: Lessons from Building a Massive German Corpus
Final Project Reports
Lessons from the Trenches of Building LLM Translation Model from Scratch
The Interplay of MT Evaluation and Multilingual LLMs
Tokenization for LLMs and Machine Translation: Morphological Plausibility, Cross-lingual Alignment, and What Actually Matters
Towards Multilingual, Multimodal Foundation Models
Tutorial: Human Evaluation of Multilingual Tasks
What Makes a Speech LLM Efficient. Metrics, Trade-offs, Design Principles, and Post-training Solutions
When Data Is the Challenge: Machine Translation for Low-resource Languages
Upload Lecture Video
Upload Lecture Video