AI RESEARCH
Voting with the Graph: Stable RLAIF via Topological Consistency Maximization
arXiv CS.AI
•
ArXi:2510.15514v3 Announce Type: replace Reinforcement Learning from AI Feedback (RLAIF) relies on LLM judges as preference measurement instruments, yet these instruments are fundamentally limited by random measurement errors -- stochastic fluctuations that manifest as preference cycles (e.g., $A \succ B \succ C \succ A$), occurring in 5-9% of evaluations across state-of-the-art models.