Skip to content

Workflows

This section provides step-by-step guides for common tasks in mi-crow. Each workflow is designed to be self-contained and complete.

Available Workflows

Saving Activations

Learn how to collect and save activations from model layers for analysis and SAE training.

When to use: - Preparing data for SAE training - Analyzing activation patterns - Debugging model behavior - Collecting datasets for research

Training SAE Models

Complete guide to training sparse autoencoders on saved activations.

When to use: - Training your first SAE - Understanding hyperparameter selection - Monitoring training progress - Saving and loading trained models

Concept Discovery

Discover interpretable concepts by analyzing SAE neuron activations.

When to use: - Finding what each SAE neuron represents - Collecting examples for manual curation - Validating concept quality - Understanding model features

Concept Manipulation

Control model behavior by manipulating discovered concepts.

When to use: - Steering model outputs - Running intervention experiments - A/B testing concept effects - Real-time model control

Activation Control

Directly manipulate activations using hooks for fine-grained control.

When to use: - Custom intervention experiments - Fine-grained activation modification - Multi-layer interventions - Advanced control patterns

Workflow Dependencies

Most workflows build on each other:

Saving Activations
    ↓
Training SAE Models
    ↓
Concept Discovery
    ↓
Concept Manipulation

Activation Control can be used independently or in combination with SAE-based workflows.

Quick Reference

Workflow Input Output Time
Saving Activations Model + Dataset Saved activations Minutes
Training SAE Saved activations Trained SAE Hours
Concept Discovery Trained SAE + Dataset Top texts per neuron Minutes
Concept Manipulation Trained SAE + Concepts Modified outputs Seconds
Activation Control Model + Hooks Modified outputs Seconds

Getting Started

If you're new to mi-crow, we recommend following workflows in order:

  1. Start with Saving Activations to understand data collection
  2. Move to Training SAE Models to learn feature discovery
  3. Try Concept Discovery to find interpretable features
  4. Use Concept Manipulation to control model behavior

For advanced users, Activation Control provides direct hook-based control.

Integration with Examples

Each workflow references relevant example notebooks:

  • Examples are in the examples/ directory
  • Workflows explain the concepts
  • Examples provide runnable code
  • See Examples Guide for the full list

Next Steps

Choose a workflow that matches your goal: