conftrace_

last-layer retraining

4 papers

Explore in graph

Also known as

LLR

Co-occurring keywords · papers mentioning both

spurious correlation (3) worst-group accuracy (2) text-based debiasing (1) linear projector (1) fair model (1) neural network (1) subgroup fairness (1) group robustness (1) feature reweighting (1) image classifier (1)

Papers

TLDR: Text Based Last-Layer Retraining for Debiasing Image Classifiers WACV 2025 Enhancing Robustness of Last Layer Two-Stage Fair Model Corrections NeurIPS 2024 Last Layer Re-Training is Sufficient for Robustness to Spurious Correlations ICLR 2023 Towards Last-layer Retraining for Group Robustness with Fewer Annotations NeurIPS 2023