Tag
该论文提出一种知识引导的LLM框架，结合语法约束语义解析与知识图谱驱动的澄清循环，用于集成地面与非地面6G网络中的意图驱动服务编排，在200条意图数据集上达到89%分类准确率与95%策略有效性。
PhysAI-Bench is a benchmark for evaluating LLM-based agentic decision-making in autonomous UAV systems, comprising over 10,000 standardized instances and evaluating 29 foundation models.
The paper proposes CRAFT, a data-centric method for fine-tuning small language models to enable pre-hoc explainability in AI-native 6G RAN, achieving higher accuracy and energy efficiency compared to existing methods like GRPO.
This survey provides a comprehensive review of wireless foundation models, covering architectures, pre-training paradigms, applications, and challenges for AI-native 6G networks.