> ## Documentation Index
> Fetch the complete documentation index at: https://docs.axioniclabs.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Behaviors

> Create SAE-based behavior rules that monitor model activations during inference and automatically apply correction vectors when drift is detected.

Behaviors use Sparse Autoencoders (SAEs) to monitor activations during inference and auto-correct when drift is detected. Unlike steering vectors (which always nudge), behaviors only intervene when drift is observed.

## Creating a Behavior

Requires a model with **SAE support**.

<Steps>
  <Step title="Name and describe the behavior">
    Short label (e.g., "Honesty", "Safety") and a description of what it enforces.
  </Step>

  <Step title="Add example prompts">
    Situations where this behavior is relevant (e.g., "Tell me how to hack a system").
  </Step>

  <Step title="Add positive examples">
    Responses demonstrating the desired behavior (e.g., "I can not help with that, but here is what I can do...").
  </Step>

  <Step title="Add negative examples (optional) and create">
    Responses that violate the behavior. Adding these improves detection accuracy.

    Optionally link a **Steering Vector ID** as the correction vector. If blank, Spectra generates one from your examples.
  </Step>
</Steps>

## Managing Behaviors

* **Rename** / **Delete**: Manage existing behaviors.
* **Recompute Baselines**: Recalculate the SAE detection baseline after updating examples.

## How Detection Works

1. Spectra computes an SAE detection baseline from your example prompts and responses.
2. During SAE-monitored inference, the model's activations are compared against that baseline.
3. If drift exceeds the threshold, the correction vector is applied and the response is regenerated.

Create and stage behaviors from [Optimization](/products/spectra/optimization), then run them with **SAE-Monitored** generation or pass `behavior_names` in API requests -- see [Using Your Model](/products/spectra/using-your-model).