无需重训模型，即可在推理时灵活调节大模型性格特征。
Controllable and explainable personality sliders for LLMs at inference time
- 通过顺序自适应引导，让性格控制向量互不干扰。
- 可同时调节五大性格维度，效果优于传统方法。
- 适合需要快速生成不同风格对话的场景。
将大语言模型（LLMs）对齐特定人格通常依赖昂贵且单一的有监督微调（SFT）或强化学习人类反馈（RLHF）。尽管有效，这些方法需为每个目标人格训练独立模型。推理时激活引导提供了一种参数高效替代方案，但朴素方法因向量干扰难以同时控制多个特质。本文提出一种模块化框架，实现连续、多维的人格控制。核心创新是顺序自适应引导（SAS）：通过在前序干预后的残差流上训练后续探测器，使引导向量正交化。该方法将引导向量变为可复用的基元，用户仅需调整系数α，即可即时合成复杂高保真人格。我们在大五人格维度上验证了该框架，结果表明其在目标遵循性和连贯性上均优于基线，实现无需更新参数的精准、整体人格调节。
原文摘要 · Abstract (English)
Aligning Large Language Models (LLMs) with specific personas typically relies on expensive and monolithic Supervised Fine-Tuning (SFT) or RLHF. While effective, these methods require training distinct models for every target personality profile. Inference-time activation steering offers a parameter-efficient alternative, yet naive approaches fail to control multiple traits simultaneously due to destructive vector interference. In this work, we propose a modular framework for continuous, multi-dimensional personality control. Our key innovation is Sequential Adaptive Steering (SAS): a method that orthogonalizes steering vectors by training subsequent probes on the residual stream shifted by prior interventions. This approach transforms steering vectors into reusable primitives, allowing users to instantly synthesize complex, high-fidelity personality profiles by simply adjusting coefficients alpha. We validate our framework on the Big Five personality traits, demonstrating that it outperforms naive baselines in both goal adherence and coherence, enabling precise, holistic personality modulation without updating model parameters.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。