English(EN) Controlling Refusal Behavior of LLMs via Stiefel-Constrained Rotation Steering

新方法增强了对LLM拒绝行为的控制

作者 PulseAugur 编辑部 · [1 个来源] · 2026-09-01 04:00

研究人员开发了一种名为Stiefel约束旋转转向的新方法，以更好地控制大型语言模型（LLM）的拒绝行为。该技术使用黎曼优化来学习模型激活参数高效的旋转变换，无需拒绝向量等辅助结构。该方法已通过实证验证，显示出提高的干预效率，并强调了特定设计选择的重要性。 AI

影响这项研究提供了一种更可靠的控制LLM输出的方法，有望提高安全性和可用性。

排序理由该集群包含一篇学术论文，详细介绍了一种控制LLM行为的新方法。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 24 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 该集群包含一篇学术论文，详细介绍了一种控制LLM行为的新方法。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: paper, safety; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: Breaking (< 6h); Fresh story with cross-source coverage still developing. Ranking may shift as more sources report.

完整方法见我们的编辑标准。

arXiv cs.CL TIER_1 English(EN) · Kirill Bunin, Dmitry Bylinkin, Vladimir Aletov, Daniil Medyakov, Vladimir Solodkin, Aleksandr Beznosikov · 2026-09-01 04:00

通过Stiefel约束旋转引导控制LLM的拒绝行为

arXiv:2608.30986v1 Announce Type: cross Abstract: Activation steering has emerged as a lightweight approach for controlling model refusal at inference time. A growing line of research explores trainable rotations of activations to develop geometrically principled intervention mec…