# Privacy-Utility Trade-off Evaluation on ChatDoctor (worst-case attack setting)

Source: [Privacy-Utility Trade-off Evaluation on ChatDoctor (worst-case attack setting)](<https://www.sota2.com/research/sota/privacy-utility-trade-off-evaluation-on-chatdoctor-worst-case-attack-setting>)

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

Data updated: 2026-04-14T10:07:48.864Z

Evaluation of the privacy-utility trade-off when running ChatDoctor models under worst-case adversarial attack conditions, measured by CRR and BERTScore.

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 | Privacy-Utility Trade-off Evaluation |
| Dataset (including any stated split) | ChatDoctor worst-case attack setting |
| Metric | CRR |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 5 of 5 |
| 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

- [CRR](<https://www.sota2.com/research/sota/privacy-utility-trade-off-evaluation-on-chatdoctor-worst-case-attack-setting.md?metric=CRR>): Higher is better; unit: PERCENT.

- [BERTScore](<https://www.sota2.com/research/sota/privacy-utility-trade-off-evaluation-on-chatdoctor-worst-case-attack-setting.md?metric=BERTScore>): Higher is better; unit: SCALAR.

## Recorded results

| Method | CRR | Result date | Report paper |
| --- | --- | --- | --- |
| No Defense | 97.4 | Not recorded | 69ddca1d7d0172140ec79a8a |
| Reranker | 95 | Not recorded | 69ddca1d7d0172140ec79a8a |
| Summarize | 76 | Not recorded | 69ddca1d7d0172140ec79a8a |
| RAGFort | 50.2 | Not recorded | 69ddca1d7d0172140ec79a8a |
| CanaryRAG | 1.6 | Not recorded | 69ddca1d7d0172140ec79a8a |

### No Defense

CRR: 97.4. Result ID: 69de11ec47f34ce924c605fb.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca1d7d0172140ec79a8a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Reranker

CRR: 95. Result ID: 69de11ec47f34ce924c605fc.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca1d7d0172140ec79a8a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa72581e209131cea55cc7.md>)

[Implementation](<https://github.com/phycholosogy/RAG-privacy>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Summarize

CRR: 76. Result ID: 69de11ec47f34ce924c605fd.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca1d7d0172140ec79a8a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa77101e209131cea70e21.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### RAGFort

CRR: 50.2. Result ID: 69de11ec47f34ce924c605fe.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca1d7d0172140ec79a8a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6916b7670052cac1dc65de15.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### CanaryRAG

CRR: 1.6. Result ID: 69de11ec47f34ce924c605ff.

[Reporting paper](<https://www.sota2.com/research/paper/69ddca1d7d0172140ec79a8a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/69ddca1d7d0172140ec79a8a.md>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

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