Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint) 文章

ArXiv CS.AI2026-05-27NEWSen作者: Paul Sigloch, Christoph Benzm\"uller
人工智能
查看原文 →
关系图谱

Neuro-Symbolic Verification of LLM Outputs for Data-Sensitive Domains (extended preprint) · 相关技术

相关技术

ginORMLLMsafetydivide-and-conquer partitioningUCTUARTEEStraight-Through EstimatorSEMSAFEPromptPPRNPUNATMITLMMHISForEvaluation Driven DevelopmentEffort Metric AttentionANNAGITableTPU