Annotation on DafnyBench 100 hardest tasks
88Pass CountIDS
Evaluation Results
| Method | Links | |
|---|---|---|
| IDSmodality=proof-only, component=DSA2026.05 | 88 | |
| Codexmodality=proof-only2026.05 | 78 | |
| Claude Codemodality=proof-only2026.05 | 76 | |
| DafnyPromodality=proof-only2026.05 | 52 |