# Software Engineering Repair on SWE Multi

Source: [Software Engineering Repair on SWE Multi](<https://www.sota2.com/research/sota/software-engineering-repair-on-swe-multi>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-05-29T11:41:10.773Z

Evaluation of model performance on software engineering repair tasks using the SWE Multi dataset, measured by SWE Avg.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | Software Engineering Repair |
| Dataset (including any stated split) | SWE Multi |
| Metric | SWE Average Score |
| Unit | SCALAR |
| Direction | Higher is better |
| Results loaded | 10 of 10 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [SWE Average Score](<https://www.sota2.com/research/sota/software-engineering-repair-on-swe-multi.md?metric=SWE%20Average%20Score>): Higher is better; unit: SCALAR.

## Recorded results

| Method | SWE Average Score | Result date | Report paper |
| --- | --- | --- | --- |
| Raw Mixture | 40 | Not recorded | 6a191dbedbd7150506f82a17 |
| MIRA-Group | 36.33 | Not recorded | 6a191dbedbd7150506f82a17 |
| Random | 35 | Not recorded | 6a191dbedbd7150506f82a17 |
| DataMan | 33 | Not recorded | 6a191dbedbd7150506f82a17 |
| MIRA-Global | 32 | Not recorded | 6a191dbedbd7150506f82a17 |
| MIRA-Source | 30.33 | Not recorded | 6a191dbedbd7150506f82a17 |
| DSIR | 27 | Not recorded | 6a191dbedbd7150506f82a17 |
| PPL | 20 | Not recorded | 6a191dbedbd7150506f82a17 |
| +SFT | 3.67 | Not recorded | 6a191dbedbd7150506f82a17 |
| Base Model | 0.33 | Not recorded | 6a191dbedbd7150506f82a17 |

### Raw Mixture

SWE Average Score: 40. Result ID: 6a197b4cfbc373592c14d3bd.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Full Corpus","Token Budget":"50B"}. This record does not assert independent reproduction.

### MIRA-Group

SWE Average Score: 36.33. Result ID: 6a197b4cfbc373592c14d3c3.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### Random

SWE Average Score: 35. Result ID: 6a197b4cfbc373592c14d3c0.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### DataMan

SWE Average Score: 33. Result ID: 6a197b4cfbc373592c14d3c1.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77d81e209131cea7585d.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### MIRA-Global

SWE Average Score: 32. Result ID: 6a197b4cfbc373592c14d3c2.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### MIRA-Source

SWE Average Score: 30.33. Result ID: 6a197b4cfbc373592c14d3c4.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### DSIR

SWE Average Score: 27. Result ID: 6a197b4cfbc373592c14d3be.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa71411e209131cea5025c.md>)

[Implementation](<https://github.com/p-lambda/dsir>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### PPL

SWE Average Score: 20. Result ID: 6a197b4cfbc373592c14d3bf.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70921e209131cea4beb4.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"Subset Selection","Token Budget":"25B"}. This record does not assert independent reproduction.

### +SFT

SWE Average Score: 3.67. Result ID: 6a197b4cfbc373592c14d3bc.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"No Mid-training","Token Budget":"0"}. This record does not assert independent reproduction.

### Base Model

SWE Average Score: 0.33. Result ID: 6a197b4cfbc373592c14d3bb.

[Reporting paper](<https://www.sota2.com/research/paper/6a191dbedbd7150506f82a17.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa75b71e209131cea6981e.md>)

Recorded evaluation configuration: {"Mid-training Protocol":"No Mid-training","Token Budget":"0"}. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.