InfraLens

ControlNet / Conditioning Reading

This starter is annotated reading material and the source of truth for the lab preview. Running anything is optional; the reading goal is to explain the mechanism without hiding behind a framework call.

Reading focus

Read ControlNet-style conditioning as an extra structural signal path into denoising.

Annotated sketch

## Control path

```text
control_image -> control_encoder -> residual features
prompt -> text_encoder -> text condition
latent, t, text condition, residual features -> denoiser

```

What to explain

• control_image is encoded separately.
• control_features align with denoiser blocks.
• The base denoiser still owns the main noise prediction.

Common trap

• ControlNet is not a negative prompt.
• Control strength is not a universal quality knob.

This site is open source. Improve this page.