Join us for an online event dedicated to exploring principles, tools and applications in the domain of Generative AI. Whether you're a beginner looking to dive into Generative AI for LLMs (Large Language Models) or for Computer Vision, this event will have something for everyone. We will discuss the role of Generative AI in today's world and the importance of various applications. We like to have online virtual events for both Large Language Models (LLMs), Models for Computer Vision and combination of such models. There will be events to discuss about evolutions and innovations of transformer-based models, solutions for model collapse and hallucinations, general reasoning and/or mathematical reasoning in Generative AI, centralized Vs decentralized models, knowledge graphs and entity-relationship based models, LLMops, Visionops etc. Our speakers will cover various important topics.
Google Meet Link: https://meet.google.com/nwm-dyeu-xdv
(Time in Pacific Daylight Time (UTC-7), California (USA) time)
12:00 PM - 12:05 PM Introduction
12:05 PM - 12: 50 PM Session 1
Speaker: Arvind Rapaka, CEO of CartUp AI
Speaker Bio: Arvind Rapaka is the founder and CEO of CartUp AI, an assisted experience platform. CartUp AI is pioneering multi-functional LLMs and building foundational technology for the next-generation eCommerce experience. Prior to founding CartUp AI, he was the Head/VP at Nextdoor, Dragonfruit, and Answers, which exited with a valuation of 1 billion USD. Additionally, he co-founded two companies (AI and eCommerce startups) that had successful exits.
Abstract: In this webinar, we explore the inner workings of Transformers in Large Language Models (LLMs) and identify key shortcomings that affect their efficiency and scalability. Despite their success, Transformers struggle with computational and memory inefficiencies, especially in handling long sequences. We will introduce the S4 model (State-Space Sequence-to-Sequence model) as a promising alternative that addresses these limitations. S4 models, leveraging continuous-time dynamical systems, offer significant improvements in processing long data sequences more efficiently. Our discussion will cover the principles behind S4, its advantages over traditional Transformers, and its potential applications in next-generation AI architectures. Join us to discover how these innovative approaches are shaping the future of artificial intelligence.
12:50 PM - 12:58 PM Q/A