MAP-PO trains one LLM per annotator cluster for sexism detection, and shows that a shared team-level reward stops agents from overshooting their cluster's labeling behavior.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Learning Sexism Detection Using Multi-Agent Perspectivist Preference Optimization
MAP-PO trains one LLM per annotator cluster for sexism detection, and shows that a shared team-level reward stops agents from overshooting their cluster's labeling behavior.