analysis: compare three CLLN scaling endpoints

1 files changed, 6 insertions, 6 deletions

diff --git a/visual-composer/coupled-ladder-scaling.md b/visual-composer/coupled-ladder-scaling.md
index c339551..896c02b 100644
--- a/visual-composer/coupled-ladder-scaling.md
+++ b/visual-composer/coupled-ladder-scaling.md
@@ -17,12 +17,13 @@
 - **Figure prototype:** coordinated small-multiple lines.
 - **Panel map:**
   - (a) final classification error against learnable edge count;
-  - (b) fraction of runs that reach zero error and remain there through the
-    training horizon;
-  - (c) mean classification error over epochs at 2,048 edges.
+  - (b) classification-error area over the full training trajectory against
+    learnable edge count;
+  - (c) fraction of runs that fail to reach zero error and remain there through
+    the training horizon against learnable edge count.
 - **Exact label inventory:** Clean, same-RMS noise, raw imperfection, static
-  calibration, SDIL, learnable edges, classification error, stable zero-error
-  runs, training epoch.
+  calibration, SDIL, learnable edges, classification error, classification-
+  error AUC, stable failure fraction.
 - **Caption role:** state the paired protocol, distinguish fixed bias from
   matched noise, and label the evidence as a five-task pilot.
 - **Manuscript placement:** Part 2, immediately after the digital CLLN and
@@ -34,4 +35,3 @@
 - **Constraint:** the figure reports an error-growth slope comparison only
   when the static-calibration slope is positive. Raw imperfection is allowed
   to show a high non-monotonic floor rather than a forced power law.
-