UnifiedQA-v2: Stronger Generalization via Broader Cross-Format Training
arXiv:2202.12359v174 citations
Originality Synthesis-oriented
AI Analysis
This work addresses the challenge of building more robust QA models for broader applications, but it is incremental as it builds directly on an existing approach.
The authors tackled the problem of improving question-answering (QA) generalization by scaling up cross-format training, resulting in better in-domain and cross-domain performance with roughly 3x more datasets than the previous model.
We present UnifiedQA-v2, a QA model built with the same process as UnifiedQA, except that it utilizes more supervision -- roughly 3x the number of datasets used for UnifiedQA. This generally leads to better in-domain and cross-domain results.