1
Probabilistic Concept-Aware Steering for Trustworthy LLM Inference
新论文提出概率概念感知引导方法，提升大语言模型推理的可信度与可控性
arXiv:2607.18259v1 Announce Type: new Abstract: Steering vectors (SVs), an inference-time intervention technique for large language models (LLMs), gui…