# Activation Steering

> Activation steering is a technique for controlling a language model's behavior at inference time by adding a steering vector, typically the difference between the model's internal activations on contrasting inputs, to its hidden activations during the forward pass, without changing its weights.

- Identifier: PTL-0219
- Category: Prompt Optimization
- Canonical URL: https://protologue.com/t/activation-steering/
- Also known as: activation addition, activation engineering, steering vectors
- Introduced: 2023

## Description

A steering vector can be computed from as little as one pair of contrasting prompts, such as "Love" and "Hate", or by averaging activation differences across many positive and negative examples, a variant called contrastive activation addition. Scaling the vector up or down, or subtracting it, strengthens or weakens the targeted behavior or topic. It needs no gradient-based optimization, but it requires access to the model's internal activations, which hosted models do not expose.

## Related terms

- [Prompt Tuning](https://protologue.com/t/prompt-tuning/)
- [Low-Rank Adaptation](https://protologue.com/t/low-rank-adaptation/)
- [Fine-Tuning](https://protologue.com/t/fine-tuning/)

## Sources

- Turner et al. (2023). Steering Language Models With Activation Engineering. https://arxiv.org/abs/2308.10248
- Panickssery et al. (2023). Steering Llama 2 via Contrastive Activation Addition. https://arxiv.org/abs/2312.06681

## Cite this entry

Protologue. (2026). Activation Steering. In Protologue: A Taxonomy of Prompting and LLM Techniques (v1.4.0, PTL-0219). https://protologue.com/t/activation-steering/

License: CC BY 4.0 (https://creativecommons.org/licenses/by/4.0/)