Please see below for the next talk in the fall seminar series organized by the Institute for AI & Computational Research on AI/ML techniques and applications across various scientific domains. You can find a table of upcoming talks here: https://web.uri.edu/iacr/seminars/.

Speaker: Marco Alvarez  (URI)

Date/Time/Location: Nov 6, 3pm, Memorial Union room 308.

Title: LLMs as Embedding Models

Abstract: Embeddings lie at the core of modern machine learning, powering applications from information retrieval to recommendation systems. In this talk, we explore the foundations and practical frontiers of embeddings derived from large language models (LLMs). We begin by examining how embeddings can be generated from pre-trained LLMs, followed by fine-tuning and adaptation techniques that align these representations with domain-specific objectives. Beyond methodology, we present preliminary results on evaluating LLM-based embeddings of source code, illustrating both their strengths and inherent limitations. Attendees will gain a grounded understanding of the current state of LLM embeddings and practical considerations for their effective use.