The paper claims that curated 20 to 40 percent subsets of training data can match full-data models and that shared label errors can inflate validation scores.
Human-Machine Collaboration on Image Annotation
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
background 1
citation-polarity summary
fields
cs.LG 1years
2025 1verdicts
REJECT 1roles
background 1polarities
unclear 1representative citing papers
citing papers explorer
-
The Achilles Heel of AI: Fundamentals of Risk-Aware Training Data for High-Consequence Models
The paper claims that curated 20 to 40 percent subsets of training data can match full-data models and that shared label errors can inflate validation scores.