gradio markdown edit

demo_watermark.py

@@ -461,6 +461,9 @@ def run_gradio(args, model=None, device=None, tokenizer=None):
                         and as the bias becomes very large the watermark transitions from "soft" to "hard". 
                         For a hard watermark, nearly all tokens are green, but this can have a detrimental effect on
                         generation quality, especially when there is not a lot of flexibility in the distribution.
+
+            #### Detector Parameters:
+            
             - z-score threshold : the z-score cuttoff for the hypothesis test. Higher thresholds (such as 4.0) make
                                 _false positives_ (predicting that human/unwatermarked text is watermarked) very unlikely
                                 as a genuine human text with a significant number of tokens will almost never achieve