Claire Gardent, CNRS, Université de Lorraine (France) Text generation from various sources into multiple languages (presentation 1, presentation 2)
Sara Hooker, Cohere (USA) Understanding the role of data, scale and capacity in recent breakthroughs (presentation)
André Martins, IT Lisboa & Unbabel (Portugal) xCOMET and Tower: Open Multilingual LLMs for Translation and Evaluation (presentation)
Barbara Plank, LMU Munich (Germany) Trustworthiness in NLP: On tackling sources of uncertainty in human label variation and models (presentation)
Thomas Wolf, Hugging Face A little guide to training LLMs in 2024 (presentation)
Maxime Peyrard, CNRS, Laboratoire d’Informatique de Grenoble, France Producing Useful Causal Explanations of LLM behavior (presentation)