# Semantic Textual Similarity on STS12

Source: [Semantic Textual Similarity on STS12](<https://www.sota2.com/research/sota/semantic-textual-similarity-on-sts12>)

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

Data updated: 2026-02-26T05:41:15.481Z

Evaluation of Semantic Textual Similarity performance on the STS12 dataset, focusing on metrics related to leakage and downstream impact.

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 | Semantic Textual Similarity |
| Dataset (including any stated split) | STS12 |
| Metric | Downstream Performance |
| 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

- [Downstream Performance](<https://www.sota2.com/research/sota/semantic-textual-similarity-on-sts12.md?metric=Downstream%20Performance>): Higher is better; unit: PERCENT.

- [Leakage](<https://www.sota2.com/research/sota/semantic-textual-similarity-on-sts12.md?metric=Leakage>): Lower is better; unit: PERCENT.

- [Tradeoff Rate R](<https://www.sota2.com/research/sota/semantic-textual-similarity-on-sts12.md?metric=Tradeoff%20Rate%20R>): Higher is better; unit: SCALAR.

## Recorded results

| Method | Downstream Performance | Result date | Report paper |
| --- | --- | --- | --- |
| Unprotected | 74.25 | Not recorded | 698abbaec97a0b3e3b5de035 |
| SPARSE | 73.25 | Not recorded | 698abbaec97a0b3e3b5de035 |
| Semantic-Replace | 64.46 | Not recorded | 698abbaec97a0b3e3b5de035 |
| Random-Replace | 60.5 | Not recorded | 698abbaec97a0b3e3b5de035 |
| RemovePII | 59.47 | Not recorded | 698abbaec97a0b3e3b5de035 |

### Unprotected

Downstream Performance: 74.25. Result ID: 6996dcfd5698665ec4297122.

[Reporting paper](<https://www.sota2.com/research/paper/698abbaec97a0b3e3b5de035.md>)

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

### SPARSE

Downstream Performance: 73.25. Result ID: 6996dcfd5698665ec4297126.

[Reporting paper](<https://www.sota2.com/research/paper/698abbaec97a0b3e3b5de035.md>)

[Method origin paper](<https://www.sota2.com/research/paper/698abbaec97a0b3e3b5de035.md>)

Recorded evaluation configuration: {"epsilon":20}. This record does not assert independent reproduction.

### Semantic-Replace

Downstream Performance: 64.46. Result ID: 6996dcfd5698665ec4297125.

[Reporting paper](<https://www.sota2.com/research/paper/698abbaec97a0b3e3b5de035.md>)

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

### Random-Replace

Downstream Performance: 60.5. Result ID: 6996dcfd5698665ec4297124.

[Reporting paper](<https://www.sota2.com/research/paper/698abbaec97a0b3e3b5de035.md>)

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

### RemovePII

Downstream Performance: 59.47. Result ID: 6996dcfd5698665ec4297123.

[Reporting paper](<https://www.sota2.com/research/paper/698abbaec97a0b3e3b5de035.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.