SEA: Personalizing Large Language Models

In this edition of SEA Single Shot, we will host a great talk by Hamed Zamani (University of Massachusetts Amherst) on personalizing Large Language Models.

***
IMPORTANT: You will be able to view the Zoom link once you 'attend' the meetup on this page.
***

Title: Personalizing Large Language Models
Time: 17:00 - 17:45 (+ 15 min Q&A), room: LAB42 L3.36
Abstract: Large Language Models (LLMs) have recently been adopted by a wide range of applications. In this talk, I will discuss models and evaluation methodologies for conditioning the LLM outputs on a user profile. In more detail, I will first introduce the Language Model Personalization (LaMP) benchmark ([https://lamp-benchmark.github.io/](https://eur04.safelinks.protection.outlook.com/?url=https%3A%2F%2Flamp-benchmark.github.io%2F&data=05%7C02%7Cp.k.hager%40uva.nl%7C0cd5eaec72964cac954008dc2c194758%7Ca0f1cacd618c4403b94576fb3d6874e5%7C0%7C0%7C638433734301082097%7CUnknown%7CTWFpbGZsb3d8eyJWIjoiMC4wLjAwMDAiLCJQIjoiV2luMzIiLCJBTiI6Ik1haWwiLCJXVCI6Mn0%3D%7C0%7C%7C%7C&sdata=qOsys8Mzs2Qvt8yg0u%2BnO9NgFAme%2FsF5BkC1jW%2FihqU%3D&reserved=0)) -- a large-scale benchmark for studying personalization for text classification and generation using LLMs. I will later draw connections between LLM personalization and retrieval-enhanced machine learning (REML) and introduce retrieval-augmented approaches for personalizing large language models. I will conclude with future directions in this area.

SEA Talk #263