# General LLM Benchmarking on ProSparse-LLaMA2-7B Evaluation Suite (test) Source: [General LLM Benchmarking on ProSparse-LLaMA2-7B Evaluation Suite (test)]() Collection index: [llms.txt]() Data updated: 2026-03-17T09:42:27.247Z Evaluation suite assessing general LLM capabilities and efficiency by aggregating results across TriviaQA, ARC-E, ARC-C, HumanEval, MBPP, GSM8K, BBH, and CMMLU, including an E2E Speedup measurement. 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 | General LLM Benchmarking | | Dataset (including any stated split) | ProSparse-LLaMA2-7B Evaluation Suite (TriviaQA, ARC-E, ARC-C, HumanEval, MBPP, GSM8K, BBH, CMMLU) (test) | | Metric | TriviaQA Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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 - [TriviaQA Accuracy](): Higher is better; unit: PERCENT. - [ARC-E Accuracy](): Higher is better; unit: PERCENT. - [ARC-C Accuracy](): Higher is better; unit: PERCENT. - [HumanEval Pass@1](): Higher is better; unit: PERCENT. - [MBPP Pass@1](): Higher is better; unit: PERCENT. - [GSM8K Accuracy](): Higher is better; unit: PERCENT. - [BBH Accuracy](): Higher is better; unit: PERCENT. - [CMMLU Accuracy](): Higher is better; unit: PERCENT. - [Average Score](): Higher is better; unit: PERCENT. - [E2E Speedup](): Higher is better; unit: SCALAR. ## Recorded results | Method | TriviaQA Accuracy | Result date | Report paper | | --- | --- | --- | --- | | Dense baseline | 62.94 | Not recorded | 69b8e0360636e5bf4afbecc9 | | SVDP (seq.) | 62.74 | Not recorded | 69b8e0360636e5bf4afbecc9 | | SVDP | 62.66 | Not recorded | 69b8e0360636e5bf4afbecc9 | | PowerInfer | 61.1 | Not recorded | 69b8e0360636e5bf4afbecc9 | | GRIFFIN | 59.77 | Not recorded | 69b8e0360636e5bf4afbecc9 | | Deja Vu | 59.47 | Not recorded | 69b8e0360636e5bf4afbecc9 | | GRIFFIN | 35.74 | Not recorded | 69b8e0360636e5bf4afbecc9 | ### Dense baseline TriviaQA Accuracy: 62.94. Result ID: 69b921f247f34ce924a22b93. [Reporting paper]() Recorded evaluation configuration: {"Predicted sparsity (%)":"−"}. This record does not assert independent reproduction. ### SVDP (seq.) TriviaQA Accuracy: 62.74. Result ID: 69b921f247f34ce924a22b99. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Predictor category":"Ours","Predicted sparsity (%)":"50","r":"256 sequential","Execution pipeline":"sequential"}. This record does not assert independent reproduction. ### SVDP TriviaQA Accuracy: 62.66. Result ID: 69b921f247f34ce924a22b98. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Predictor category":"Ours","Predicted sparsity (%)":"75","r":"1024 parallel"}. This record does not assert independent reproduction. ### PowerInfer TriviaQA Accuracy: 61.1. Result ID: 69b921f247f34ce924a22b95. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Predictor category":"Trained predictors","Predicted sparsity (%)":"80","r":"1024 parallel"}. This record does not assert independent reproduction. ### GRIFFIN TriviaQA Accuracy: 59.77. Result ID: 69b921f247f34ce924a22b96. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Predictor category":"Training-free methods","Predicted sparsity (%)":"20","r":"1024 parallel"}. This record does not assert independent reproduction. ### Deja Vu TriviaQA Accuracy: 59.47. Result ID: 69b921f247f34ce924a22b94. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Predictor category":"Trained predictors","Predicted sparsity (%)":"20","r":"1024 parallel"}. This record does not assert independent reproduction. ### GRIFFIN TriviaQA Accuracy: 35.74. Result ID: 69b921f247f34ce924a22b97. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Predictor category":"Training-free methods","Predicted sparsity (%)":"50","r":"1024 parallel"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.