Outlier-Aware Training for Improving Group Accuracy Disparities

2022-10-27 · Li-Kuang Chen, Canasai Kruengkrai, Junichi Yamagishi

Methods addressing spurious correlations such as Just Train Twice (JTT, arXiv:2107.09044v2) involve reweighting a subset of the training set to maximize the worst-group accuracy. However, the reweighted set of examples may potentially contain unlearnable examples that hamper the model's learning. We propose mitigating this by detecting outliers to the training set and removing them before reweighting. Our experiments show that our method achieves competitive or better accuracy compared with JTT and can detect and remove annotation errors in the subset being reweighted in JTT.

📄 PDF Abstract BibTeX arXiv:2210.15183

Code (0)

등록된 구현이 없습니다.

Centering the Margins: Outlier-Based Identification of Harmed Populations in Toxicity Detection

2023-05-24 · Vyoma Raman, Eve Fleisig, Dan Klein

The impact of AI models on marginalized communities has traditionally been measured by identifying performance differences between specified demographic subgroups. Though this approach aims to center vulnerable groups, i…

Outlier Detection

Fairness-Aware Fine-Tuning of Vision-Language Models for Medical Glaucoma Diagnosis

2025-12-03 · Zijian Gu, Yuxi Liu, Zhenhao Zhang, Song Wang arxiv

Vision-language models achieve expert-level performance on medical imaging tasks but exhibit significant diagnostic accuracy disparities across demographic groups. We introduce fairness-aware Low-Rank Adaptation for medi…

Mitigating Individual Skin Tone Bias in Skin Lesion Classification through Distribution-Aware Reweighting

2025-12-09 · Kuniko Paxton, Zeinab Dehghani, Koorosh Aslansefat, Dhavalkumar Thakker 외 arxiv

Skin color has historically been a focal point of discrimination, yet fairness research in machine learning for medical imaging often relies on coarse subgroup categories, overlooking individual-level variations. Such gr…

Skin Lesion Classification Density Estimation

Benchmarking Bias Mitigation Toward Fairness Without Harm from Vision to LVLMs

2026-02-03 · Xuwei Tan, Ziyu Hu, Xueru Zhang arxiv

Machine learning models trained on real-world data often inherit and amplify biases against certain social groups, raising urgent concerns about their deployment at scale. While numerous bias mitigation methods have been…

Q-ROAR: Outlier-Aware Rescaling for RoPE Position Interpolation in Quantized Long-Context LLMs

2025-09-17 · Ye Qiao, Sitao Huang arxiv

Extending LLM context windows is crucial for long range tasks. RoPE-based position interpolation (PI) methods like linear and frequency-aware scaling extend input lengths without retraining, while post-training quantizat…