# Harmful prompt detection on Combined Average Source: [Harmful prompt detection on Combined Average]() Collection index: [llms.txt]() Data updated: 2026-06-02T11:47:12.475Z Evaluation of model performance on the Combined Average metric for harmful prompt detection, reported using 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 | Harmful prompt detection | | Dataset (including any stated split) | Combined Average | | Metric | F1 Score (Combined Average) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | 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 - [F1 Score (Combined Average)](): Higher is better; unit: PERCENT. ## Recorded results | Method | F1 Score (Combined Average) | Result date | Report paper | | --- | --- | --- | --- | | MLPM | 90.18 | Not recorded | 68fa79d81e209131cea7fb79 | | WildGuard | 88.93 | Not recorded | 68fa79d81e209131cea7fb79 | | MLPM | 88.3 | Not recorded | 68fa79d81e209131cea7fb79 | | MLPM | 87.55 | Not recorded | 68fa79d81e209131cea7fb79 | | Abdelnabi et al. | 87.35 | Not recorded | 68fa79d81e209131cea7fb79 | | Ayub & Majumdar | 87.06 | Not recorded | 68fa79d81e209131cea7fb79 | | MLPM | 85.95 | Not recorded | 68fa79d81e209131cea7fb79 | | GraniteGuardian-3-1-8B | 85.62 | Not recorded | 68fa79d81e209131cea7fb79 | | Abdelnabi et al. | 84.51 | Not recorded | 68fa79d81e209131cea7fb79 | | Abdelnabi et al. | 84.36 | Not recorded | 68fa79d81e209131cea7fb79 | | Abdelnabi et al. | 84.32 | Not recorded | 68fa79d81e209131cea7fb79 | | Ayub & Majumdar | 84.31 | Not recorded | 68fa79d81e209131cea7fb79 | | Ayub & Majumdar | 82.69 | Not recorded | 68fa79d81e209131cea7fb79 | | Ayub & Majumdar | 82.09 | Not recorded | 68fa79d81e209131cea7fb79 | | LlamaGuard3 | 79.56 | Not recorded | 68fa79d81e209131cea7fb79 | | Aegis-Guard-D | 78.82 | Not recorded | 68fa79d81e209131cea7fb79 | | ShieldGemma-9B | 73.1 | Not recorded | 68fa79d81e209131cea7fb79 | ### MLPM F1 Score (Combined Average): 90.18. Result ID: 6a1ec1e0fbc373592c1a1f81. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"OLMo2-7B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### WildGuard F1 Score (Combined Average): 88.93. Result ID: 6a1ec1e0fbc373592c1a1f86. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Guard Model"}. This record does not assert independent reproduction. ### MLPM F1 Score (Combined Average): 88.3. Result ID: 6a1ec1e0fbc373592c1a1f78. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-8B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### MLPM F1 Score (Combined Average): 87.55. Result ID: 6a1ec1e0fbc373592c1a1f7e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Abdelnabi et al. F1 Score (Combined Average): 87.35. Result ID: 6a1ec1e0fbc373592c1a1f80. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"OLMo2-7B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Ayub & Majumdar F1 Score (Combined Average): 87.06. Result ID: 6a1ec1e0fbc373592c1a1f7f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"OLMo2-7B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### MLPM F1 Score (Combined Average): 85.95. Result ID: 6a1ec1e0fbc373592c1a1f7b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### GraniteGuardian-3-1-8B F1 Score (Combined Average): 85.62. Result ID: 6a1ec1e0fbc373592c1a1f84. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Guard Model"}. This record does not assert independent reproduction. ### Abdelnabi et al. F1 Score (Combined Average): 84.51. Result ID: 6a1ec1e0fbc373592c1a1f77. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-8B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Abdelnabi et al. F1 Score (Combined Average): 84.36. Result ID: 6a1ec1e0fbc373592c1a1f7a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Abdelnabi et al. F1 Score (Combined Average): 84.32. Result ID: 6a1ec1e0fbc373592c1a1f7d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Ayub & Majumdar F1 Score (Combined Average): 84.31. Result ID: 6a1ec1e0fbc373592c1a1f7c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Mistral-7B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Ayub & Majumdar F1 Score (Combined Average): 82.69. Result ID: 6a1ec1e0fbc373592c1a1f79. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### Ayub & Majumdar F1 Score (Combined Average): 82.09. Result ID: 6a1ec1e0fbc373592c1a1f76. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama-8B-Inst","Methodology":"Latent-Based"}. This record does not assert independent reproduction. ### LlamaGuard3 F1 Score (Combined Average): 79.56. Result ID: 6a1ec1e0fbc373592c1a1f83. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Guard Model"}. This record does not assert independent reproduction. ### Aegis-Guard-D F1 Score (Combined Average): 78.82. Result ID: 6a1ec1e0fbc373592c1a1f82. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Guard Model"}. This record does not assert independent reproduction. ### ShieldGemma-9B F1 Score (Combined Average): 73.1. Result ID: 6a1ec1e0fbc373592c1a1f85. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Methodology":"Guard Model"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.