Practical GenAI Problems in the Context of Information Theory

Paul Gesiak, Director, Data Analytics Consulting at EPAM, gave a presentation at EPAM’s annual event — Learning Week 2023. Anywhere Club is now making the video* of Paul’s presentation available to registered members of the Club.

Practical GenAI Problems in the Context of Information Theory
This video is only available to Anywhere Club members

A brief description of Paul’s presentation: 

We are witnessing the rapid evolution of Generative AI that is ongoing and will continue. Focusing on models such as GPT-4, Claude 2 and LLaMa2, Paul explores the main challenges of using these models, including data privacy, bias, and misinformation risks. Key unknowns in AI, like explainability and unpredictability of behavior, are also discussed. Practical applications are shown in model training, with entropy helping to prevent model collapse and improve algorithm efficiency. Further research to enhance the understanding and functionality of Generative AI models is clearly not a whim, but a necessity. 

* The following was recorded at an EPAM Systems, Inc. event and contains information which, at the time of its presentation, was confidential. EPAM Systems, Inc. is the sole owner of this video and its contents, including, without limitation, any presentations or information presented. This video contains proprietary information and is being delivered to you for the specific purpose, and subject to the terms and conditions, as agreed between you and EPAM Systems, Inc. This video is protected under the copyright laws of the United States and other countries, and any unauthorized duplication, distribution or exhibition of this video or its contents may result in civil liability and criminal prosecution.
Related posts
Get the latest updates on the platforms you love