DeepRCL/SteeringGeometry

Tasks

• text-classification

Citation text from the card

@inproceedings{abootorabi-etal-2026-steering,
    title = "Steering Geometry: Validating Human Value Geometry in {LLM} Steering Space",
    author = "Abootorabi, Mohammad Mahdi  and
      Saghafian, Armin  and
      Bazshoushtari, Ali  and
      Rezaei, Hamid  and
      Hwang, EunJeong  and
      Shwartz, Vered  and
      Mousavi, Parvin  and
      Abolmaesumi, Purang",
    editor = "Lee, Hung-Yi  and
      Dev, Sunipa  and
      Li, Jessy  and
      Elliott, Desmond",
    booktitle = "Proceedings of the 2026 Conference on Empirical Methods in Natural Language Processing",
    month = oct,
    year = "2026",
    address = "Budapest, Hungary",
    publisher = "Association for Computational Linguistics",
    url = "https://aclanthology.org/2026.emnlp-main.2508/",
    pages = "57010--57033",
    ISBN = "979-8-89176-459-0",
    abstract = "As large language models (LLMs) are increasingly deployed in alignment-sensitive contexts, activation steering has emerged as a lightweight, inference-time alternative to fine-tuning methods (e.g., RLHF, DPO) for behavioral control. However, existing work typically validates steering on isolated behaviors, leaving it unclear whether steering vectors encode coherent semantic structure or merely exploit behavior-specific shortcuts. We investigate whether the latent geometry of LLM steering vectors reflects theory-specified structure in human values and morality. Using Schwartz{'}s Theory of Basic Human Values as our primary fine-grained framework

All datasets marked Unclear | JSON