The Ghost Annotator framework applies conformal prediction and collaborative filtering representations to measure LLM divergence from human annotations across four models and datasets, revealing higher confidence in misaligned cases and consistent demographic misalignment.
Are you sure? Measuring models bias in content moderation through uncertainty
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
The Ghost Annotator: a Framework to Explore Human Label Variation in Content Moderation through Conformal Prediction
The Ghost Annotator framework applies conformal prediction and collaborative filtering representations to measure LLM divergence from human annotations across four models and datasets, revealing higher confidence in misaligned cases and consistent demographic misalignment.