LLMs achieve 64% accuracy detecting Wikipedia bias and remove 79% of words removed by editors when correcting, but produce high-recall low-precision edits rated more neutral by crowds than human versions.
Human-Machine Collaboration for Content Regulation: The Case of Reddit Automoderator
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2024 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Seeing Like an AI: How LLMs Apply (and Misapply) Wikipedia Neutrality Norms
LLMs achieve 64% accuracy detecting Wikipedia bias and remove 79% of words removed by editors when correcting, but produce high-recall low-precision edits rated more neutral by crowds than human versions.