LMMs for AI-Native Wireless Systems - AI/ML webinar

#AIML #futurenetworks #webinar #ingr

Special Presentation on “LMMs as Universal Foundation Models for AI-Native Wireless Systems”

by Dr. Christo K. Thomas (Virginia Tech, USA)

Hosted by Future Networks Artificial Intelligence & Machine Learning (AIML) Working Group

Date/Time: Thursday, March 7^th, 2024 @ 6 PM EST (3 PM PST)

Topic:

Large Multi-Modal Models (LMMs) as Universal Foundation Models for AI-Native Wireless Systems

Abstract:

Foundation models such as large language models (LLMs) have recently been touted as game-changers for 6G systems. However, previous efforts on LLMs for wireless networks are limited to directly applying existing language models designed for natural language processing (NLP) applications. Contrary to this, in this talk, we present a comprehensive vision of how to design universal foundation models that are tailored towards the unique needs of next-generation wireless systems, thereby paving the way towards the deployment of artificial intelligence (AI)-native networks. These LMMs are driven by three distinct characteristics: 1) integration of multi-modal sensing data, 2) grounding sensory input via causal reasoning and retrieval-augmented generation (RAG), and 3) instructibility to environmental feedback through logical and mathematical reasoning enabled by neuro-symbolic AI. These attributes are crucial for developing "universal foundation models" capable of addressing interconnected cross-layer networking challenges in AI-native wireless systems while ensuring alignment of objectives across diverse domains. We also discuss preliminary results from experimental evaluation that demonstrate the efficacy of grounding using RAG in LMMs, and showcase the alignment of LMMs with wireless system designs. Furthermore, compared to vanilla LLMs, the enhanced rationale exhibited in the responses to mathematical questions by LMMs demonstrates the logical and mathematical reasoning capabilities inherent in LMMs. Building on those results, we present a sequel of open questions and challenges for LMMs, including intent-based networks, resilient wireless systems, semantic communications, and many more.