MaChAmp at SemEval-2023 Tasks 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, and 12: On the Effectiveness of Intermediate Training on an Uncurated Collection of Datasets

Computer Science

• Natural Language Processing 100%
• Processing Task 50%
• Performance Gain 50%
• Language Modeling 50%
• Intermediate Step 50%
• Performance Improvement 50%
• Pre-Trained Language Models 50%

Keyphrases

• Transfer of Rights 50%
• Substantial Performance 50%
• Performance Gains 50%
• Transfer Tasks 50%