BARRED: Synthetic Training of Custom Policy Guardrails via Asymmetric Debate

arXiv:2604.25203v1 Announce Type: cross Abstract: Deploying guardrails for custom policies remains challenging, as generic safety models fail to capture task-specific requirements, while prompting LLMs suffers from inconsistent boundary-case performance and high inference costs. Training custom classifiers achieves both accuracy and efficiency, yet demands substantial labeled data that is costly to obtain. We present BARRED (Boundary Alignment Refinement through REflection and Debate), a framewo