The Defeat of the Winograd Schema Challenge
Data Skeptic
00:00
Advancements in Natural Language Processing and Pre-trained Models
This chapter explores the progress made in natural language processing, particularly in relation to the Winograd Schema Challenge. The speaker discusses their work in common sense reasoning, pre-training in large language models, and knowledge base completion, highlighting the use of pre-trained models like BERT and early GPTs.
Transcript
Play full episode
