Toward Stable Value Alignment: Introducing Independent Modules for Consistent Value Guidance

arXiv:2605.11712v1 Announce Type: new Abstract: Aligning large language models (LLMs) with human values typically relies on post-training or inference-time steering that directly manipulates the backbone's parameters or representation space. However, a critical gap exists: the model's residual stream is highly dynamic, in which values exist as fragile, low-dimensional properties, inherently incompatible with the stability required for consistent value expression. In this paper, we propose the St