# Privacy-utility tradeoff on STS12 Source: [Privacy-utility tradeoff on STS12]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:48:33.848Z Evaluation of the trade-off between privacy preservation and data utility using the STS12 dataset, measured by Leakage, Confidence, and Downstream performance. 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 tradeoff | | Dataset (including any stated split) | STS12 | | Metric | Leakage | | Unit | SCALAR | | Direction | Lower is better | | Results loaded | 16 of 16 | | 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 - [Leakage](): Lower is better; unit: SCALAR. - [Confidence](): Higher is better; unit: SCALAR. - [Downstream Performance](): Higher is better; unit: SCALAR. ## Recorded results | Method | Leakage | Result date | Report paper | | --- | --- | --- | --- | | SPARSE | 4.34 | Not recorded | 698abbaec97a0b3e3b5de035 | | LapMech | 7.36 | Not recorded | 698abbaec97a0b3e3b5de035 | | PurMech | 7.42 | Not recorded | 698abbaec97a0b3e3b5de035 | | SPARSE | 19.31 | Not recorded | 698abbaec97a0b3e3b5de035 | | LapMech | 22.34 | Not recorded | 698abbaec97a0b3e3b5de035 | | PurMech | 22.66 | Not recorded | 698abbaec97a0b3e3b5de035 | | SPARSE | 36.98 | Not recorded | 698abbaec97a0b3e3b5de035 | | PurMech | 38.04 | Not recorded | 698abbaec97a0b3e3b5de035 | | LapMech | 38.17 | Not recorded | 698abbaec97a0b3e3b5de035 | | SPARSE | 43.81 | Not recorded | 698abbaec97a0b3e3b5de035 | | LapMech | 44.74 | Not recorded | 698abbaec97a0b3e3b5de035 | | PurMech | 44.76 | Not recorded | 698abbaec97a0b3e3b5de035 | | SPARSE | 47.54 | Not recorded | 698abbaec97a0b3e3b5de035 | | PurMech | 48.34 | Not recorded | 698abbaec97a0b3e3b5de035 | | LapMech | 48.48 | Not recorded | 698abbaec97a0b3e3b5de035 | | Unprotected embedding | 60.09 | Not recorded | 698abbaec97a0b3e3b5de035 | ### SPARSE Leakage: 4.34. Result ID: 6996dcfd5698665ec4297129. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"5"}. This record does not assert independent reproduction. ### LapMech Leakage: 7.36. Result ID: 6996dcfd5698665ec4297127. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"5"}. This record does not assert independent reproduction. ### PurMech Leakage: 7.42. Result ID: 6996dcfd5698665ec4297128. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"5"}. This record does not assert independent reproduction. ### SPARSE Leakage: 19.31. Result ID: 6996dcfd5698665ec429712c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"10"}. This record does not assert independent reproduction. ### LapMech Leakage: 22.34. Result ID: 6996dcfd5698665ec429712a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"10"}. This record does not assert independent reproduction. ### PurMech Leakage: 22.66. Result ID: 6996dcfd5698665ec429712b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"10"}. This record does not assert independent reproduction. ### SPARSE Leakage: 36.98. Result ID: 6996dcfd5698665ec429712f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"20"}. This record does not assert independent reproduction. ### PurMech Leakage: 38.04. Result ID: 6996dcfd5698665ec429712e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"20"}. This record does not assert independent reproduction. ### LapMech Leakage: 38.17. Result ID: 6996dcfd5698665ec429712d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"20"}. This record does not assert independent reproduction. ### SPARSE Leakage: 43.81. Result ID: 6996dcfd5698665ec4297132. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"30"}. This record does not assert independent reproduction. ### LapMech Leakage: 44.74. Result ID: 6996dcfd5698665ec4297130. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"30"}. This record does not assert independent reproduction. ### PurMech Leakage: 44.76. Result ID: 6996dcfd5698665ec4297131. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"30"}. This record does not assert independent reproduction. ### SPARSE Leakage: 47.54. Result ID: 6996dcfd5698665ec4297135. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"40"}. This record does not assert independent reproduction. ### PurMech Leakage: 48.34. Result ID: 6996dcfd5698665ec4297134. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"40"}. This record does not assert independent reproduction. ### LapMech Leakage: 48.48. Result ID: 6996dcfd5698665ec4297133. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"epsilon":"40"}. This record does not assert independent reproduction. ### Unprotected embedding Leakage: 60.09. Result ID: 6996dcfd5698665ec4297136. [Reporting paper]() Recorded evaluation configuration: {"epsilon":"infinity"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.