English

COLD-Steer: Steering Large Language Models via In-Context One-step Learning Dynamics

Machine Learning 2026-03-09 v1 Artificial Intelligence Computation and Language

Abstract

Activation steering methods enable inference-time control of large language model (LLM) behavior without retraining, but current approaches face a fundamental trade-off: sample-efficient methods suboptimally capture steering signals from labeled examples, while methods that better extract these signals require hundreds to thousands of examples. We introduce COLD-Steer, a training-free framework that steers LLM activations by approximating the representational changes that would result from gradient descent on in-context examples. Our key insight is that the effect of fine-tuning on a small set of examples can be efficiently approximated at inference time without actual parameter updates. We formalize this through two complementary approaches: (i) a unit kernel approximation method that updates the activations directly using gradients with respect to them, normalized across examples, and (ii) a finite-difference approximation requiring only two forward passes regardless of example count. Experiments across a variety of steering tasks and benchmarks demonstrate that COLD-Steer achieves upto 95% steering effectiveness while using 50 times fewer samples compared to the best baseline. COLD-Steer facilitates accommodating diverse perspectives without extensive demonstration data, which we validate through our experiments on pluralistic alignment tasks. Our framework opens new possibilities for adaptive, context-aware model control that can flexibly address varying loss-driven human preferences through principled approximation of learning dynamics rather than specialized training procedures.

Keywords

Cite

@article{arxiv.2603.06495,
  title  = {COLD-Steer: Steering Large Language Models via In-Context One-step Learning Dynamics},
  author = {Kartik Sharma and Rakshit S. Trivedi},
  journal= {arXiv preprint arXiv:2603.06495},
  year   = {2026}
}

Comments

ICLR 2026. Code available at https://github.com/Ksartik/cold-steer

R2 v1 2026-07-01T11:07:20.095Z