# Functional Correctness on HumanEval (p@k and Precision) Source: [Functional Correctness on HumanEval (p@k and Precision)]() Collection index: [llms.txt]() Data updated: 2026-06-05T09:51:39.529Z Evaluation of model performance on the HumanEval dataset for functional correctness, reporting Pass@k metrics (k=1, 10, 100) and Precision metrics. 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 | Functional Correctness | | Dataset (including any stated split) | HumanEval | | Metric | Pass@1 | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 27 of 27 | | 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 - [Pass@1](): Higher is better; unit: PERCENT. - [Pass@10](): Higher is better; unit: PERCENT. - [Pass@100](): Higher is better; unit: PERCENT. - [Precision](): Higher is better; unit: PERCENT. ## Recorded results | Method | Pass@1 | Result date | Report paper | | --- | --- | --- | --- | | Dense | 71.57 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 54.79 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 46.43 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 41.95 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 38.16 | Not recorded | 6a2258ae86173ad9c1468c67 | | Dense | 36.7 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 34.37 | Not recorded | 6a2258ae86173ad9c1468c67 | | Dense | 27.67 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 26.2 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 21.37 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 21.3 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 18.94 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 18.7 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 16.1 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 15.32 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 15.06 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 14.96 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 14.16 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 11.91 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 11.7 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 10.55 | Not recorded | 6a2258ae86173ad9c1468c67 | | SecRL-Prune | 8.75 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 7.98 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 7.9 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 5.96 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 5.8 | Not recorded | 6a2258ae86173ad9c1468c67 | | PruneNet | 4.2 | Not recorded | 6a2258ae86173ad9c1468c67 | ### Dense Pass@1: 71.57. Result ID: 6a229b59fbc373592c212df6. [Reporting paper]() Recorded evaluation configuration: {"Compression":"0%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 54.79. Result ID: 6a229b59fbc373592c212df9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"10%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 46.43. Result ID: 6a229b59fbc373592c212dfa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"10%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 41.95. Result ID: 6a229b59fbc373592c212dff. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"20%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 38.16. Result ID: 6a229b59fbc373592c212e05. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"25%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### Dense Pass@1: 36.7. Result ID: 6a229b59fbc373592c212df8. [Reporting paper]() Recorded evaluation configuration: {"Compression":"0%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 34.37. Result ID: 6a229b59fbc373592c212e0b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"30%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### Dense Pass@1: 27.67. Result ID: 6a229b59fbc373592c212df7. [Reporting paper]() Recorded evaluation configuration: {"Compression":"0%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 26.2. Result ID: 6a229b59fbc373592c212dfd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"10%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 21.37. Result ID: 6a229b59fbc373592c212dfb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"10%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 21.3. Result ID: 6a229b59fbc373592c212e00. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"20%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 18.94. Result ID: 6a229b59fbc373592c212e03. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"20%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 18.7. Result ID: 6a229b59fbc373592c212e06. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"25%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 16.1. Result ID: 6a229b59fbc373592c212e0c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"30%","Backbone":"Qwen2.5-Coder-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 15.32. Result ID: 6a229b59fbc373592c212e09. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"25%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 15.06. Result ID: 6a229b59fbc373592c212e01. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"20%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 14.96. Result ID: 6a229b59fbc373592c212dfc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"10%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 14.16. Result ID: 6a229b59fbc373592c212dfe. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"10%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 11.91. Result ID: 6a229b59fbc373592c212e07. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"25%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 11.7. Result ID: 6a229b59fbc373592c212e0f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"30%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 10.55. Result ID: 6a229b59fbc373592c212e04. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"20%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### SecRL-Prune Pass@1: 8.75. Result ID: 6a229b59fbc373592c212e0d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"30%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 7.98. Result ID: 6a229b59fbc373592c212e02. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"20%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 7.9. Result ID: 6a229b59fbc373592c212e0a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"25%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 5.96. Result ID: 6a229b59fbc373592c212e08. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"25%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 5.8. Result ID: 6a229b59fbc373592c212e10. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"30%","Backbone":"CodeLlama-7B-Python"}. This record does not assert independent reproduction. ### PruneNet Pass@1: 4.2. Result ID: 6a229b59fbc373592c212e0e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Compression":"30%","Backbone":"CodeLlama-7B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.