T 0.32 Y
Detected: injected error lowered target-dim F1 (Y = detected)
T 1.0 X
Missed: target-dim F1 stayed 1.0 despite injection (X = not detected)
C 1.0 =
GT≡Gen: LLM found no modifiable element, gen code = gt code, F1 = 1.0
FAIL
Compile Failed: .tex was generated but could not compile to SVG
-
Not Generated: no error variant was produced for this dimension
ERR
Evaluation Error: runtime error during F1 evaluation
Each card shows GT + 4 error variants (Type/Text/BBox/Color). Click image tabs to compare. The F1 bar breakdown highlights the injected dimension with a yellow arrow. Click any image to zoom.