Researcher Calls Data Quality Highest-Leverage Factor in AI Training
Jonathan Schwarz of Thomson Reuters and Imperial College shares insight on foundation models.
Jonathan Richard Schwarz, Head of AI Research at Thomson Reuters and Visiting Professor at Imperial College London, stated that data quality stands out as both the most underappreciated and highest-leverage element in foundation model performance. Everyone familiar with the training process recognizes this priority, according to his post. The remark was shared via a link to adaption_ai. The packet contains only this attributed quote with no additional corroboration or outcomes reported.
Combined views
1.6K
1 post, first seen 3h ago
Researcher Calls Data Quality Highest-Leverage Factor in AI Training
Jonathan Schwarz of Thomson Reuters and Imperial College shares insight on foundation models.
Jonathan Richard Schwarz, Head of AI Research at Thomson Reuters and Visiting Professor at Imperial College London, stated that data quality stands out as both the most underappreciated and highest-leverage element in foundation model performance. Everyone familiar with the training process recognizes this priority, according to his post. The remark was shared via a link to adaption_ai. The packet contains only this attributed quote with no additional corroboration or outcomes reported.
Reactions
1 postJR 
Jonathan Richard Schwarz@schwarzjn_Everyone who has worked on foundation model training knows this: data quality is often the most underappreciated - and simultaneously highest-leverage - factor in model performance.
- likes: 16
- replies: 2
- bookmarks: 7
- reposts: 0