О нас
The University of San Francisco Data Science & Artificial Intelligence Speaker Series is produced by the Data Institute. This group brings researchers and practitioners together with students in the MS in Data Science and Artificial Intelligence graduate program, faculty, and interested members of the public to discuss topics of interest in analytics and data science.
Talks take place in-person on Fridays from 12:30-2:00 pm at the USF Campus in Downtown San Francisco, located at 101 Howard Street. You may view the schedule below and we encourage you to bring your lunch.
Talk recordings are made available subject to speaker permission. You can find the recorded talks at https://www.youtube.com/channel/UCN0kf0sI01-FXPZdWAA-uMA
Предстоящие мероприятия (1)
Смотреть все- From Word2Vec to GPT: How AI Learns Language and How We Evaluate Them101 Howard St, University of San Francisco - Downtown Campus, San Francisco, CA 94105, San Francisco, CA
We are excited to welcome back our alumna Viviana Márquez. Viviana is the AI Community Lead at Prolific, where she helps AI teams build better machine learning systems by providing fast, self-serve access to diverse human data across the AI lifecycle. She holds a Master’s in Data Science from the University of San Francisco and has worked as a data scientist at HBO, Dataminr, and Royal Caribbean, among others. As an educator, she has trained more than 4,000 learners worldwide, from Fortune 500 teams to graduate students, transforming complex AI concepts into practical skills.
Fun fact: In 2021, Viviana competed in the Miss Universe Colombia contest as Miss Sucre, placing in the top 13.Description
How do machines learn the meaning of words, and how do we know if they’ve learned it well? In this talk, we will explore the evolution of natural language processing, beginning with early representations such as one-hot vectors, moving through the breakthrough of Word2Vec, and continuing to the rise of transformers and today’s generative models like GPT. Along the way, we will examine the differences between representation models and generative models, and why evaluation is straightforward for one but a major challenge for the other.
The session will then dive into modern evaluation methods, including code-based checks and LLMs-as-judges, with a special focus on human-in-the-loop evaluation. We will explore how real human judgment is used to measure quality, safety, and usefulness in language AI. Attendees will leave with both a historical perspective and a practical framework for thinking about evaluation in their own projects.🔗 Register now to join us.
#USFDataScience #SpeakerSeries #NaturalLanguageProcessing #GenerativeAI #MachineLearning #ArtificialIntelligence #AIethics #DataScience
