Program Repair 벤치마크
Program Repair on DeepFix
Average Success Rate
- 2018-01-31 — RLAssist: Average Success Rate 26.6
- 2019-06-24 — SampleFix: Average Success Rate 45.3
- 2020-05-20 — DrRepair: Average Success Rate 68.2
- 2021-06-11 — DrRepair + BIFI: Average Success Rate 71.7
| Rank | Model | Average Success Rate | Paper | Code | Year |
|---|---|---|---|---|---|
| 1 | DrRepair + BIFI | 71.7 | Break-It-Fix-It: Unsupervised Learning for Program Repair | michiyasunaga/bifi | 2021 |
| 2 | DrRepair | 68.2 | Graph-based, Self-Supervised Program Repair from Diagnostic Feedback | michiyasunaga/DrRepair · worksheets/0x01838644 | 2020 |
| 3 | SampleFix | 45.3 | SampleFix: Learning to Generate Functionally Diverse Fixes | 2019 | |
| 4 | RLAssist | 26.6 | Deep Reinforcement Learning for Programming Language Correction | terne/dtuproject | 2018 |
| 5 | DrRepair + BIFI | 71.7 | Break-It-Fix-It: Unsupervised Learning for Program Repair | michiyasunaga/bifi | 2021 |
| 6 | DrRepair | 68.2 | Graph-based, Self-Supervised Program Repair from Diagnostic Feedback | michiyasunaga/DrRepair · worksheets/0x01838644 | 2020 |
| 7 | SampleFix | 45.3 | SampleFix: Learning to Generate Functionally Diverse Fixes | 2019 | |
| 8 | RLAssist | 26.6 | Deep Reinforcement Learning for Programming Language Correction | terne/dtuproject | 2018 |