Why Steering Works: Toward a Unified View of Language Model Parameter Dynamics

ACL ARR 2026 January Submission2059 Authors

Keywords: LLM control; steering; representation engineering; LoRA; dynamic weight updates; preference–utility trade-off; representation manifold

Abstract: Methods for controlling large language models (LLMs), including local weight fine-tuning, LoRA-based adaptation, and activation-based interventions, are often studied in isolation, obscuring their connections and making comparison difficult. In this work, we present a unified view that frames these interventions as dynamic weight updates induced by a control signal, placing them within a single conceptual framework. Building on this view, we propose a unified preference-utility analysis that separates control effects into preference, defined as the tendency toward a target concept, and utility, defined as coherent and task-valid generation, and measures both on a shared log-odds scale using polarity-paired contrastive examples. Across methods, we observe a consistent trade-off between preference and utility: stronger control increases preference while predictably reducing utility. We further explain this behavior through an activation manifold perspective, in which control shifts representations along target-concept directions to enhance preference, while utility declines primarily when interventions push representations off the model’s valid-generation manifold. Finally, we introduce a new steering approach guided by this analysis that improves preference while better preserving utility.

Paper Type: Long

Research Area: Safety and Alignment in LLMs

Research Area Keywords: model editing, probing, robustness

Contribution Types: Model analysis & interpretability

Languages Studied: English

Submission Number: 2059