Published Jun 1, 2017
12 - Supervised Learning of Universal Sentence Representations from Natural Language Inference Data
Explore the complexities of training universal sentence representations using the SNLI dataset, as experts Waleed Ammar and Matt Gardner unravel the benefits and challenges of supervised learning methods in natural language processing. Delve into the nuances of dataset evaluation and transfer learning challenges that shape the limitations and potential of current sentence encoding models.
Topics covered
Popular Clips
Episode Highlights
Related Episodes
38 - A Corpus of Natural Language for Visual Reasoning, with Alane Suhr
Answers 383 questions09 - Learning to Generate Reviews and Discovering Sentiment
Answers 383 questions13 - Question Answering from Unstructured Text by Retrieval and Comprehension
Answers 383 questions
11 - Relation Extraction with Matrix Factorization and Universal Schemas
Answers 383 questions
48 - Incidental Supervision: Moving Beyond Supervised Learning, with Dan Roth
Answers 383 questions
64 - Neural Network Models for Sentence Pair Tasks, with Wuwei Lan and Wei Xu
Answers 383 questions25 - Neural Semantic Parsing over Multiple Knowledge-bases
Answers 383 questions20 - A simple neural network module for relational reasoning
Answers 383 questions
116 - Grounded Language Understanding, with Yonatan Bisk
Answers 383 questions01 - A Comparative Study of Word Embeddings for Reading Comprehension
Answers 383 questions28 - Data Programming: Creating Large Training Sets, Quickly
Answers 383 questions
