Different Corruptions, Different Signals: Uncertainty and Loss in Federated Data Quality
Federated learning (FL) data corruption can affect either inputs or labels, but it remains unclear whether input-conditional uncertainty and prediction-label loss expose these corruption modes equally. This paper compares two corruption-detection signals in FL: input-conditional uncertainty and prediction-label loss. The uncertainty signal is characterised using a learned aleatoric variance…
We haven't written up this one. arXiv cs.AI has the full story — the link below goes straight to it.