Work

Slice-based Learning: A Programming Model for Residual Learning in Critical Data Slices

Year: 2019
Type: article
Abstract: predictions. We show that our approach maintains a parameter-efficient representation while improving over baselines by up to 19.0 F1 on slices and 4.6 F1 overall on datasets spanning language understanding (e.g. SuperGLUE), computer vision, and production-scale industrial systems.
Source: PubMed
Institution: Stanford University
Language: English
Open Access status: green
OpenAlex ID: W2971350322
FWCI: 3.32
Cites: 41
Cited by: 52
Related to: 10