Update README.md

README.md

@@ -197,7 +197,7 @@ artistically weird things like inside Stable Diffusion, the latent space accordi
    conceptual texts, educative brochures, narratives, etc.
 
 ## Can I use to provide an abstract image reference for AI in the article or book about AI I'm writing?
-Yes, you can **as long as you cite this page and autor even if this page goes down, which is me Robb-0.** No worry about my real name or any other referrence, if I change my mind 
+Yes, you can **as long as you cite this page or author even if this page goes down, which is me Robb-0.** Don't worry about my real name or any other referrence, if I change my mind 
 I'll know how to link the new me with this lovely Robb-O here.
 
 **Attribution and License:** This model is the creative brainchild of Robb-0, licensed under the Creative Commons Attribution 4.0 International (CC BY 4.0). 
@@ -206,23 +206,23 @@ When using this model, ensure proper attribution is given to the artist.
 ### As seen on samples, results not necessarily render as expected. Still they keep the style in pace. Happens that there are at least concepts here - all intertwinted: 
 
 1. It aids to create Apple-like Avatars but in a more fluffly neon fashion
-2. It adds the idea of The Artist, which is the model per se
-3. And also the bringing the representation of the idea of a latent space to IllustriousXL, while other models seems to grasp it quite will such as SD 1.5, SDXL, and even Flux.
+2. It adds the idea of The Artist, which is the model per se. Sd and SDXL tend to answer quite well if you start the prompt with "Artist".
+3. And also the abstract concept of a latent space to IllustriousXL. Note that other models seem to grasp it quite will such as SD 1.5, SDXL, and even Flux. (Update) It's actually
+4. a representation of the Cosine wave. Once I asked for Kalman Kernels, and it delivered that circle thing. Asking for "cosine wave", "fourier wave" and the likes may help creating
+5. abstract images related to AI, which can be used just as an illustration paired alongside related text.
 
 ### Languages suggested are those which IllustriousXL and NoobAI XL seem to work best. If you have a version which supports multi-languages...
 
-Then you will be able to use them, although these "trigger words" are only in English. Am not sure if models can translate them and it will be useful in any other language.
+Saying that about the whole prompt. SDXL seems to understand more languages. I'm not sure if Illustrious XL and NoobAI XL still keep the SDXL dual-CLIP.
 
 ## Trigger words
 
-the use `trigger words are not necessary` for general image generation.
+ `trigger words are not necessary` for general image generation.
 
 You don't need to use those words, `but each reinforces some different concepts:` `latent space` to trigger the inner model view of the latent space and adapt the conecpt to your 
 image generation. There are time it adds many psichological cues, and esoteric stuff. Good for art, research,  and semiotic analysis as well as hermeneutics to see how the AI thinks.
 
 You should use `the artist` to whether summon the model to picture as the "The Artist". 
-**EVEN WITHOUT THIS LORA** You can also start the prompt with a hidden Uber-SUPER-OP quality "opening" such as 
-"dear Artist create an image of...", it works even for NoobleAI V-Pred. Note: I don't have the image I used that "dear Artist please create..." anylonger, but I created  dozeon of good quality just by "speaking" with V-Pred.
 
 You should use `latent space` to trigger the latent space element all diffusers now about it. It can be used for fun or research.
 
@@ -235,13 +235,13 @@ It's likely it can be used with another themes and LoRas for extra effects.
 ## Compatibilities
 
 The **The Artist / Style** image lora model was trained on Indigo furry Mix XL Noob AI EPS III. 
-Reason: It manages well both Dan/Gel-booru tags as natural language.
+Reason: It manages well both Dan/Gel-booru tags as well as natural language.
 
-- NoobAI EPS III model and finetunes: should be able to create nice images with this LoRA models (I tested and it works).
+- NoobAI EPS III model and finetunes: should be able to create nice images with this LoRA models (I tested and it works, samples pinned above).
 - I also have tested it on Illustrious 1, 1.1, 2 and 2 PRO. Results are dilluted. It's a hit or miss. Although they are more recent than EPS III, which may explain it.
-- Previous models such as the base Illustrious V0.1 STILL should be able to create images here. You may or may need to increase CFG and increase the steps - or switch sampler to SDE++ (loads of artifacts)
-- Pony: It's not impossible that Illustrious Loras run on Pony. I have not tested but it might work. Maybe setting weight higher than 1 may do or not.
-- SDXL base: It's likely not work, or just spree a fragrance on the images. I have not tested it either, yet odds are it won't work.
+- Previous models such as the base Illustrious V0.1 STILL should be able to create images as in here. You may or may need to increase CFG and increase the steps - or switch sampler to SDE++ (loads of artifacts)
+- Pony: It's not impossible that Illustrious Loras run on Pony. Especially when they are overfitting. This one does NOT work with Pony when I tested it. And chances are that Pony is able to render these things without any LoRA.
+- SDXL base: It's likely not to work. I have not tested it either, yet odds are it won't work.
 
 # WARNING
 
@@ -262,7 +262,7 @@ Weights for this model are available in Safetensors format.
 
 This is an image LoRA model, you need access to a UI which runs Illustrious or NoobAI models and a slot for image lora. ComfyUI and Automatic1111 (among others) to run locally.
 **OR**
-Simple visit Civitai and the Lora is available for use on site as well: https://civitai.com/models/1370366
+Simply visit Civitai and the Lora is available for use on site as well: https://civitai.com/models/1370366
 
 ---