September 9 – Pre-Day: Efficient Generative AI Summit
September 10-12, 2024 – AI Hardware & Edge AI Summit
Signia by Hilton, San Jose, CA
Rapid adoption of large language models (LLMs) has created a critical need for sharing these models while preventing the users from copying, or reverse engineering, the models or the data they represent. Fully Homomorphic Encryption (FHE) provides a post-quantum approach to secure the publication of LLMs. Recent software and hardware developments will provide a way to perform this secure sharing at execution speeds comparable to inference done on unsecured LLMs. Dr. Rhines will review the emerging approaches and provide a roadmap for future developments that will make secure data sharing of LLMs pervasive.
To see the full schedule for this series, or to register to attend: https://aihwedgesummit.com/events/aihwedgesummit