Update explanation in readme

README.md

@@ -166,9 +166,9 @@ $$\text{Normalized Loss} = \frac{Ny \times \left(Nx - \frac{Ny - 1}{2}\right) -
 The lower the loss, the closer the model is to the perfect recall curve,
 indicating higher performance.
 
-![Recall plot illustrating loss metric](https://github.com/asreview/asreview-insights/blob/main/figures/loss_metric_example.png)
+![Recall plot illustrating loss metric](https://github.com/jteijema/asreview-insights/blob/loss-metric/figures/loss_metric_example.png?raw=true)
 
-In this figure, the green area between the recall curve and the perfect recall line, represents the model's performance loss. This loss is normalized relative to the total area (green and red combined).
+In this figure, the green area between the recall curve and the perfect recall line is the lossed performance, which is then normalized for the total area (green and red combined).
 
 ## Basic usage