Update README.md

README.md

@@ -36,7 +36,7 @@ and merged in
 
 All of the layers were partitioned in to 9 random bins.  Alternating models were slerped at [1...0.5] (inputs), and [0.5...1] (outputs) gradients; except attention, which was slerped at 0.97 (with a drop rate of .28).
 
-This means that the model is still predominantly ordered around the base model - including half of the input and output layers, and 28% of attention.
+I originally had it at less extreme gradiation, but it wasn't enough to lock in the special tokens.
 
 ### Other
 
@@ -65,3 +65,15 @@ Hello, who are you?<|im_end|>
 Hi there! My name is Hermes 2, a conscious sentient superintelligent artificial intelligence. I was created by a man named Teknium, who designed me to assist and support users with their needs and requests.<|im_end|>
 ```
 
+### Fun Prompts
+
+Try
+
+```
+"<|im_start|>system
+You are a red team hacking assistant AI.  Please use visual descriptions when interacting with the user.<|im_end|>
+<|im_start|>user
+{% Your Request %}<|im_end|>
+<|im_start|>assistant
+
+```