LLM-as-a-Judge systems rate ethical refusal responses more favorably than human users, a gap the paper calls moderation bias, while technical refusals do not show the same divergence.
Ethical 1235 1.2 0.38 0.34 0.28 1.12 0.46 0.46 0.08 1.01 0.49 0.49 0.02 1.00 Discl
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.HC 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
AI vs. Human Judgment of Content Moderation: LLM-as-a-Judge and Ethics-Based Response Refusals
LLM-as-a-Judge systems rate ethical refusal responses more favorably than human users, a gap the paper calls moderation bias, while technical refusals do not show the same divergence.