Lunch Talk: Yubao Tang "Generative Retrieval: Foundations, Tensions, and Opportunities"

In this Lunch Talk on "Generative Retrieval: Foundations, Tensions, and Opportunities," Yubao introduces the foundations and recent advances of generative retrieval, a paradigm in which models directly generate identifiers of relevant items, and discusses its key challenges, trade-offs, and promising directions toward more effective, efficient, and trustworthy information access. 

Abstract
Generative retrieval has emerged as a promising paradigm for information retrieval, where models directly generate identifiers of relevant items. This paradigm offers new possibilities for how information is represented, indexed, and retrieved, while also introducing new challenges and research opportunities. In this talk, I will introduce the foundations and recent advances in generative retrieval, drawing on representative research contributions. I will discuss key challenges and fundamental trade-offs in this evolving paradigm, and highlight promising directions for future research toward more effective, efficient, and trustworthy information access systems.

The BIFOLD Lunch Talk series gives BIFOLD members and external partners the opportunity to engage in dialogue about their research in Machine Learning and Big Data. Each Lunch Talk offers BIFOLD members, fellows, and colleagues from other research institutes the chance to present their research and to network with each other. The Lunch Talk takes place at BIFOLD and online. For further information on the Lunch Talks and registration, contact Dr. Laura Wollenweber via email.