CLOct 14, 2025
On the Interplay between Human Label Variation and Model Fairness
arXiv:2510.12036v1h-index: 36
Originality Incremental advance
AI Analysis
This addresses fairness in machine learning models for applications with human-labeled data, but it appears incremental as it builds on existing HLV methods.
The paper tackled the problem of how human label variation affects model fairness, finding that training with HLV methods improves fairness without explicit debiasing.
The impact of human label variation (HLV) on model fairness is an unexplored topic. This paper examines the interplay by comparing training on majority-vote labels with a range of HLV methods. Our experiments show that without explicit debiasing, HLV training methods have a positive impact on fairness.