Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.
Rethinking soft labels for knowledge distillation: A bias-variance tradeoff perspective
4 Pith papers cite this work. Polarity classification is still indexing.
representative citing papers
SAAD adaptively weights adversarial training samples by their transferability to the teacher, yielding higher AutoAttack robustness than prior distillation methods on CIFAR and Tiny-ImageNet without extra compute.
Leakage-aware distillation transfers at least 90% of tabular foundation model AUC to lightweight students across 19 health datasets, with 26x CPU speedup and preserved calibration/fairness.
Audits miscalibration in LLM-based social science measurements across 14 constructs and proposes a soft label distillation pipeline that reduces ECE by 43.2% and Brier score by 34.0% on average.
citing papers explorer
-
Toward Calibrated, Fair, and accurate Deepfake Detection
Face-Feature Tuning is a label-free logit remapping method that reduces FPR/TPR gaps across groups in deepfake detection while preserving overall accuracy.
-
Distilling Tabular Foundation Models for Structured Health Data
Leakage-aware distillation transfers at least 90% of tabular foundation model AUC to lightweight students across 19 health datasets, with 26x CPU speedup and preserved calibration/fairness.
-
Assessing and Mitigating Miscalibration in LLM-Based Social Science Measurement
Audits miscalibration in LLM-based social science measurements across 14 constructs and proposes a soft label distillation pipeline that reduces ECE by 43.2% and Brier score by 34.0% on average.