Card: link the Civitai workflow page, add the figures

README.md

@@ -22,10 +22,22 @@ A LoRA for FLUX 3 Image Klein that nudges a restore toward your photo. When FLUX
 tiles, faces come out a little closer to the picture you gave it, and their colours stay closer. It is optional in my
 upscale workflow: without it the workflow still runs.
 
+Also on Civitai: my [FLUX 3 Image Klein Upscale](https://civitai.com/models/3007200) workflow, which uses this LoRA.
 
 > **The difference is small.** Side by side, the faces with the LoRA had slightly better colour and shape, and a
 > face-recognition check agrees: below.
 
+![A woman with black bangs, from a 386 px wide JPEG. Without the LoRA her skin comes out a little darker and her cheeks lose some pink; with it the colour stays close to the photo](figure_bangs.jpg)
+
+*In each picture: the small input, the result without the LoRA and with it, and the marked face enlarged underneath.
+All made with the workflow in this repo, Photo at 4 megapixels, same request and seed. "Without the LoRA" is the same
+workflow with the LoRA row switched off. "Same person" and "colour off" are the face check below, against the clean
+picture the input was made from. None of these pictures are in the training set.*
+
+![A woman with a bob and a black coat, from a 463 px wide JPEG. Without the LoRA her face comes out darker than in the photo; with it the face stays closer to the photo's light](figure_bob_coat.jpg)
+
+![A woman with headphones, from a 317 px wide JPEG. Neither result brings back her freckles, which the small input does not hold; with the LoRA the skin tone sits a little closer to the photo](figure_headphones.jpg)
+
 ## How to use it
 
 Put it on the **distilled** model at strength **1.0**. It was trained on the base model. The trigger is the sentence it
@@ -40,6 +52,8 @@ Reference Multi Latent**, resolution `keep resolution`), and the KSampler starts
 8 steps, CFG 1, `euler` / `simple`, denoise 1. The workflow in this repo does the whole chain: SeedVR2 3B enlarges a
 photo first, then FLUX 3 redraws it in tiles.
 
+![How the upscale workflow uses the LoRA. Pick Photo (SeedVR2 3B cleans it up and enlarges it), Blurry photo (FLUX 3 draws the whole picture again at 2 megapixels) or Generated picture (4x-UltraSharpV2 enlarges it). Then FLUX 3 redraws it in tiles of about 2 megapixels and the tiles are joined. The restore LoRA goes on FLUX 3 wherever it draws, when its file is there](figure_workflow.jpg)
+
 ## What I measured
 
 My upscale workflow's example pictures, with the LoRA on and off and everything else the same: the same request, seed
@@ -92,7 +106,8 @@ license click: [black-forest-labs/flux-3-image-klein-comfy](https://huggingface.
 [`workflows/flux-3-image-klein-upscale.json`](workflows/flux-3-image-klein-upscale.json) has it set up. Load your
 picture, pick **Photo**, **Blurry photo** or **Generated picture**, and press Run. The LoRA row is optional: without
 the file the workflow skips it and runs. **Photo** runs SeedVR2 3B first, **Generated picture** runs 4x-UltraSharpV2
-first, and the workflow's note lists every file it needs with its link.
+first, and the workflow's note lists every file it needs with its link. The same workflow is on
+[Civitai](https://civitai.com/models/3007200), inside its example pictures too.
 
 To add it to your own graph: