Baseline System Overview
James explains their baseline system, consisting of evidence retrieval and classification components. The system retrieves relevant evidence using TF IDF, selects supporting or refuting sentences, and applies a decomposable attention classifier for classification. Notably, the model handles cases with not enough evidence by training a three-way classifier.In this clip
From this podcast

NLP Highlights
60 - FEVER: a large-scale dataset for Fact Extraction and VERification, with James Thorne
Related Questions