| Dataset Name | SOTA Method | Metric | Trend | ||
|---|---|---|---|---|---|
| Real misspecified programs neutral briefs | Gemini-2.5-flash | Repair Success Rate100 | 60 | 24d ago | |
| SpreadsheetBench (test) | IG | Pass Rate84 | 18 | 2mo ago | |
| Cangjie-bench | ILA-agent | ACC90.63 | 12 | 4mo ago | |
| DeepFix (test) | BIFI | Accuracy71.7 | 9 | 4mo ago | |
| Bugs2Fix (medium) | ERNIE-Code | BLEU-491.2 | 7 | 4mo ago | |
| Bugs2Fix (small) | ERNIE-Code | BLEU-480.1 | 7 | 4mo ago | |
| Code-Contests (test) | Assistive Decomp | Strict Accuracy12.5 | 4 | 4mo ago | |
| APPS (test) | Assistive Decomp | Strict Accuracy21.7 | 4 | 4mo ago |