Only 58 of 314 eligible deep learning benchmark faults meet all four realism conditions, and only 86 of 165 reproduction attempts succeed, suggesting most 'real' DL fault benchmarks are not faithful to their sources.
Bugs in machine learning-based systems: a faultload benchmark,
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
citation-role summary
dataset 1
citation-polarity summary
fields
cs.SE 1years
2024 1verdicts
CONDITIONAL 1roles
dataset 1polarities
use dataset 1representative citing papers
citing papers explorer
-
Real Faults in Deep Learning Fault Benchmarks: How Real Are They?
Only 58 of 314 eligible deep learning benchmark faults meet all four realism conditions, and only 86 of 165 reproduction attempts succeed, suggesting most 'real' DL fault benchmarks are not faithful to their sources.