# Byzantine Client Detection on CIFAR-10 (test) Source: [Byzantine Client Detection on CIFAR-10 (test)]() Collection index: [llms.txt]() Data updated: 2026-05-19T07:22:55.195Z Evaluation of a model's ability to detect Byzantine clients using the standard CIFAR-10 test set, measured by Average Detection Accuracy and Average F1 Score. 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 | Byzantine Client Detection | | Dataset (including any stated split) | CIFAR-10 (test) | | Metric | Average Detection Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 26 of 26 | | 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 - [Average Detection Accuracy](): Higher is better; unit: PERCENT. - [Average F1 Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Average Detection Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 100 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 97.78 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 96.44 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 96 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 90.67 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 87.11 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 86.67 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 76 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 75.56 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 75.11 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 73.33 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 73.33 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 72.44 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 70.67 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 70.22 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 69.33 | Not recorded | 6a0bee88dbc836b87d82c388 | | Classical MultiKrum | 61.33 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 60 | Not recorded | 6a0bee88dbc836b87d82c388 | | QUBO-based client selection | 60 | Not recorded | 6a0bee88dbc836b87d82c388 | ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b758. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Gaussian Noise","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b75c. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Scale","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b75e. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Targeted","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b760. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Clustered","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b762. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Same Value","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b766. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Blatant Lie","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b76c. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Label Flip","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 100. Result ID: 6a0c0fb1fbc373592c01b770. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Stealthy","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 97.78. Result ID: 6a0c0fb1fbc373592c01b769. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"ALIE","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 96.44. Result ID: 6a0c0fb1fbc373592c01b75f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Targeted","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 96. Result ID: 6a0c0fb1fbc373592c01b771. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Stealthy","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 90.67. Result ID: 6a0c0fb1fbc373592c01b76f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Shuffle","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 87.11. Result ID: 6a0c0fb1fbc373592c01b75d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Scale","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 86.67. Result ID: 6a0c0fb1fbc373592c01b763. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Same Value","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 76. Result ID: 6a0c0fb1fbc373592c01b764. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Lie","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 75.56. Result ID: 6a0c0fb1fbc373592c01b768. [Reporting paper]() Recorded evaluation configuration: {"Attack":"ALIE","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 75.11. Result ID: 6a0c0fb1fbc373592c01b761. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Clustered","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 73.33. Result ID: 6a0c0fb1fbc373592c01b765. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Lie","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 73.33. Result ID: 6a0c0fb1fbc373592c01b76b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Sparse Lie","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 72.44. Result ID: 6a0c0fb1fbc373592c01b76a. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Sparse Lie","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 70.67. Result ID: 6a0c0fb1fbc373592c01b75b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Sign Flip","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 70.22. Result ID: 6a0c0fb1fbc373592c01b76e. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Shuffle","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 69.33. Result ID: 6a0c0fb1fbc373592c01b767. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Blatant Lie","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### Classical MultiKrum Average Detection Accuracy: 61.33. Result ID: 6a0c0fb1fbc373592c01b75a. [Reporting paper]() Recorded evaluation configuration: {"Attack":"Sign Flip","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 60. Result ID: 6a0c0fb1fbc373592c01b759. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Gaussian Noise","Evaluation Rounds":"30"}. This record does not assert independent reproduction. ### QUBO-based client selection Average Detection Accuracy: 60. Result ID: 6a0c0fb1fbc373592c01b76d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Attack":"Label Flip","Evaluation Rounds":"30"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.