Recent studies have indicated that Large Language Models (LLMs) harbor an inherent understanding of truthfulness, yet often fail to express fully and generate false statements. This gap between "knowing" and "telling" poses a challenge for ensuring the truthfulness of generated content. To address this, we introduce Adaptive Activation Steering (ACT), a tuning-free method that adaptively shift LLM's activations in "truthful" direction during inference. ACT addresses diverse categories of hallucinations by utilizing diverse steering vectors and adjusting the steering intensity adaptively. Applied as an add-on across various models, ACT significantly improves truthfulness in LLaMA ($\uparrow$ 142\%), LLaMA2 ($\uparrow$ 24\%), Alpaca ($\uparrow$ 36\%), Vicuna ($\uparrow$ 28\%), and LLaMA2-Chat ($\uparrow$ 19\%). Furthermore, we verify ACT's scalability across larger models (13B, 33B, 65B), underscoring the adaptability of ACT to large-scale language models.

使用适应性激活引导（ACT）方法，可以通过调整大型语言模型（LLMs）的激活方向，在推理过程中提高生成内容的真实性。ACT方法通过利用多样化的引导向量并自适应地调整引导强度，有效应对各种类型的虚构现象。在多种模型中附加应用ACT后，如LLaMA，LLaMA2，Alpaca，Vicuna和LLaMA2-Chat，其真实性显著提高（分别提升142％，24％，36％，28％和19％）。此外，我们还验证了ACT在更大规模模型（13B，33B，65B）上的可扩展性，突显ACT对大型语言模型的适应性。

自适应激活导向：无需调整的多种幻觉类别LLM真实性改进方法