Cohere For AI - Guest Speaker: Weijia Shi, PhD student at UW

Introduction: I am a PhD student in Computer Science at the University of Washington advised by Prof. Luke Zettlemoyer and Prof. Noah A. Smith. I have been a visiting research at Meta AI, working with Scott Yih. Prior to UW, I graduated from UCLA with a B.S. in Computer Science and Minor in Math. Twitter: https://twitter.com/WeijiaShi2

Abstract: Large language models (LMs) are currently trained to predict tokens given document prefixes, enabling them to directly perform long-form generation and prompting-style tasks which can be reduced to document completion. Existing pretraining pipelines train LMs by concatenating random sets of short documents to create input contexts but the prior documents provide no signal for predicting the next document. We instead present In-Context Pretraining, a new approach where language models are pretrained on a sequence of related documents, thereby explicitly encouraging them to read and reason across document boundaries. We can do In-Context Pretraining by simply changing the document ordering so that each context contains related documents, and directly applying existing pretraining pipelines. However, this document sorting problem is challenging. There are billions of documents and we would like the sort to maximize contextual similarity for every document without repeating any data. To do this, we introduce approximate algorithms for finding related documents with efficient nearest neighbor search and constructing coherent input contexts with a graph traversal algorithm.abs: https://arxiv.org/abs/2310.10638

Add event to calendar

Apple Google Office 365 Outlook Outlook.com Yahoo