Results: claude-opus-5_baseline_20260820-194627

⬇ Download submission ZIP
CADGenBench / run
Benchmark score: 0.463 Validity rate: 95.1% (77/81) Invalid: 3 Missing: 1 81 samples
generation: 0.351 editing: 0.635

Click a card to view details. j/k to navigate, Esc to return. Each card shows the input and the candidate output. Score tint: ≥0.90 ≥0.60 <0.60 CAD score.

Generation 49
Editing 32