用高斯调度调节激活值，让大模型更诚实说话
Depth-Wise Activation Steering for Honest Language Models
- 按深度分层加权激活值，用高斯分布分配干预强度
- 在七款模型上提升诚实度，六款优于基线
- 无需微调，适合想控制模型说真话的研究者
大型语言模型有时会声称错误信息，尽管内部已存储正确答案——这是诚实性而非准确性的失败，影响可审计性和安全性。现有方法多聚焦事实正确性或依赖微调与脆弱的单层编辑，对真实陈述的控制力有限。本文提出一种无需训练的激活值引导方法，通过高斯调度在模型深度上分配干预强度。在分离诚实性与知识的MASK基准测试中，评估了涵盖LLaMA、Qwen和Mistral系列的七款模型，发现高斯调度在六款模型上优于无引导和单层基线。对LLaMA-3.1-8B-Instruct与Qwen-2.5-7B-Instruct的等预算对比显示，高斯调度优于随机、均匀及盒滤波分配方式，表明干预在深度上的分布方式显著影响结果，而不仅取决于总强度。该方法简单、模型无关，无需微调，为利用模型现有能力实现诚实输出提供低成本控制手段。
原文摘要 · Abstract (English)
Large language models sometimes assert falsehoods despite internally representing the correct answer, failures of honesty rather than accuracy, which undermines auditability and safety. Existing approaches largely optimize factual correctness or depend on retraining and brittle single-layer edits, offering limited leverage over truthful reporting. We present a training-free activation steering method that weights steering strength across network depth using a Gaussian schedule. On the MASK benchmark, which separates honesty from knowledge, we evaluate seven models spanning the LLaMA, Qwen, and Mistral families and find that Gaussian scheduling improves honesty over no-steering and single-layer baselines in six of seven models. Equal-budget ablations on LLaMA-3.1-8B-Instruct and Qwen-2.5-7B-Instruct show the Gaussian schedule outperforms random, uniform, and box-filter depth allocations, indicating that how intervention is distributed across depth materially affects outcomes beyond total strength. The method is simple, model-agnostic, requires no finetuning, and provides a low-cost control knob for eliciting truthful reporting from models' existing capabilities.
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。