Research map / Label-efficient and robust learning / Recognition and long-tail
Stronger Than You Think: Benchmarking Weak Supervision on Realistic Tasks
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Benchmarks weak supervision (WS) across realistic tasks, varying data sources, labeling functions, label models, and end-model pipelines, to test its practical value.
Problem
Weak supervision is hard to benchmark because of many setup choices (data sources, labeling functions, label models, end models), so its practical value is unclear.
Approach
A benchmark of realistic WS tasks that systematically varies the WS pipeline knobs; the provided text is only the NeurIPS template and does not describe the method details.
Closest papers by meaning
- AutoWS-Bench-101: Benchmarking Automated Weak Supervision with 100 Labels (NeurIPS 2022 · similarity 0.93)
- WRENCH: A Comprehensive Benchmark for Weak Supervision (NeurIPS 2021 · similarity 0.93)
- Characterizing the Impacts of Semi-supervised Learning for Weak Supervision (NeurIPS 2023 · similarity 0.92)
- Learning with Partial-Label and Unlabeled Data: A Uniform Treatment for Supervision Redundancy and Insufficiency (ICML 2024 · similarity 0.89)
- Weak Supervision Performance Evaluation via Partial Identification (NeurIPS 2024 · similarity 0.88)
- Theoretical Analysis of Weak-to-Strong Generalization (NeurIPS 2024 · similarity 0.86)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.