All events
Workshops & Seminars

AI Open Source Day

24/06/2024
9h-14h30

Station F

Registration

**** AGENDA **** 

Doors open at 09:00 for registration and coffee.

General Track | Master Stage

  • 09:45 – 10:45

Welcome Address

Laurent Solly, Vice President Southern Europe at Meta • Roxanne Varza, Director of Station F

  • 10:15 – 11:00
  • Why Open Science and Open Source matter in AI?

Yann LeCun, VP & Chief AI Scientist at Meta • Thomas Wolf, Co-founder and CSO at Hugging Face • Patrick Perez, CEO at Kyutai . Moderated by Adrienne Jan, Chief Product Officer at Scaleway

  • 11:00 – 11:45

Meta Llama 3: from Research to Applications

Joe Spisak, Product Director & Head of Generative AI Open Source at Meta • Lysandre Debut, Head of Open Source at Hugging Face. 

Presented by DINUM (Direction Interministérielle du Numérique)

  • 11:45 – 12:30

Startups & Open Source Research Demos

  • 12:30 – 13:30

Lunch Break

  • 13:30 – 14:00

DINOv2: From Research to Real Life Application

Camille Couprie, Research Scientist at Meta • Benjamin Lefaudeux, Head of Computer Vision & AI at PhotoRoom

  • 14:00 – 14:45

How Openness Advances AI Safety

Jean-Marie John Mathews, Co-founder at Giskard AI • Guillaume Avrin, National coordinator for AI • Ania Kaci, Responsible AI Leader @ IBM France
Moderated by Pascale Fung, Senior Director of AI Research at Meta 

  • 14:45 – 15:00

Closing Remarks

Academic Track | Central Room

  • 10:00 – 10:30

Code Llama a state-of-the-art large language model for coding

Fabian Gloeckle, PHD Student at Meta

  • 10:30 – 11:00

AI Watermarking

Pierre Fernandez, PHD Student at Meta

  • 11:00 – 11:30

Introducing SeamlessM4T

Marta R. Costa-Jussa, Research Scientist at Meta

  • 11:30 – 12:00

Video Joint Embedding Predictive Architecture (V-JEPA)

Adrien Bardes, Research Scientist at Meta

  • 12:00 – 13:30

Lunch Break

  • 13:30 – 14:00

DINOv2: Learning Robust Visual Features Without Supervision

Timothée Darcet, PHD Student at Meta

  • 14:00 – 14:30

What Matters Most When Building Vision-Language Models?

Leo Tronchon, ML Research Engineer at HuggingFace