activevision commited on
Commit
3cee49a
·
verified ·
1 Parent(s): 6f16668

Update README.md

Browse files
Files changed (1) hide show
  1. README.md +0 -11
README.md CHANGED
@@ -70,17 +70,6 @@ The pipeline is the artifact. Every benchmark image is produced in two determini
70
 
71
  The released benchmark contains only the Stage-2 images. Held-out splits with unpublished seeds and additional difficulties can be regenerated from the included generators.
72
 
73
- ## Anti-shortcut design
74
-
75
- Every task is constructed to defeat six shortcut classes (full discussion in `code/scope.md`):
76
-
77
- 1. One-shot symbolic perception, then text reasoning.
78
- 2. End-to-end CV scripting (OCR / edge / blob / vector tracing).
79
- 3. Prior leakage / answer-distribution memorisation.
80
- 4. Gestalt heuristics that bypass step-by-step tracing.
81
- 5. Memorisation of the released benchmark (mitigated because the *pipeline* is released, not just the data).
82
- 6. Tool-using zoom/crop/OCR loops.
83
-
84
  ## Responsible AI
85
 
86
  See `croissant.json` for the full RAI block. Headlines:
 
70
 
71
  The released benchmark contains only the Stage-2 images. Held-out splits with unpublished seeds and additional difficulties can be regenerated from the included generators.
72
 
 
 
 
 
 
 
 
 
 
 
 
73
  ## Responsible AI
74
 
75
  See `croissant.json` for the full RAI block. Headlines: