# Zero-shot Commonsense Reasoning Suite (PIQA, SIQA, HellaSwag, ARC-E, ARC-C, WinoGrande, LAMBADA) Source: [Zero-shot Commonsense Reasoning Suite (PIQA, SIQA, HellaSwag, ARC-E, ARC-C, WinoGrande, LAMBADA)]() Collection index: [llms.txt]() Data updated: 2026-05-12T17:04:38.657Z Evaluation of zero-shot performance across seven core commonsense reasoning benchmarks: PIQA, SIQA, HellaSwag, ARC-Easy, ARC-Challenge, WinoGrande, and LAMBADA. 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 | Commonsense Reasoning | | Dataset (including any stated split) | Commonsense Reasoning Suite (PIQA, SIQA, HellaSwag, ARC-E, ARC-C, WinoGrande, LAMBADA) zero-shot | | Metric | PIQA Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 18 of 18 | | 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 - [PIQA Accuracy](): Higher is better; unit: PERCENT. - [SIQA Accuracy](): Higher is better; unit: PERCENT. - [HellaSwag Accuracy](): Higher is better; unit: PERCENT. - [ARC-E Accuracy](): Higher is better; unit: PERCENT. - [ARC-C Accuracy](): Higher is better; unit: PERCENT. - [WinoGrande Accuracy](): Higher is better; unit: PERCENT. - [LAMBADA Accuracy](): Higher is better; unit: PERCENT. - [Overall Average Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | PIQA Accuracy | Result date | Report paper | | --- | --- | --- | --- | | ReMoE | 66.92 | Not recorded | 6a02b4b4dbc836b87d824fcd | | Dense | 66.87 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DeepSeek-V3 | 66.32 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DECO | 66.14 | Not recorded | 6a02b4b4dbc836b87d824fcd | | BlockFFN | 66.1 | Not recorded | 6a02b4b4dbc836b87d824fcd | | ReMoE | 66.05 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DeepSeek-V3 | 65.94 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DECO | 65.94 | Not recorded | 6a02b4b4dbc836b87d824fcd | | TopP | 64.53 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DECO | 60.94 | Not recorded | 6a02b4b4dbc836b87d824fcd | | BlockFFN | 60.83 | Not recorded | 6a02b4b4dbc836b87d824fcd | | Dense | 60.77 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DeepSeek-V3 | 60.66 | Not recorded | 6a02b4b4dbc836b87d824fcd | | ReMoE | 60.34 | Not recorded | 6a02b4b4dbc836b87d824fcd | | TopP | 60.17 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DeepSeek-V3 | 60.12 | Not recorded | 6a02b4b4dbc836b87d824fcd | | ReMoE | 59.74 | Not recorded | 6a02b4b4dbc836b87d824fcd | | DECO | 58.22 | Not recorded | 6a02b4b4dbc836b87d824fcd | ### ReMoE PIQA Accuracy: 66.92. Result ID: 6a035d8b47f34ce924e906bc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large","Routing Strategy":"NG"}. This record does not assert independent reproduction. ### Dense PIQA Accuracy: 66.87. Result ID: 6a035d8b47f34ce924e906bf. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Large"}. This record does not assert independent reproduction. ### DeepSeek-V3 PIQA Accuracy: 66.32. Result ID: 6a035d8b47f34ce924e906b9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large","Routing Strategy":"GA"}. This record does not assert independent reproduction. ### DECO PIQA Accuracy: 66.14. Result ID: 6a035d8b47f34ce924e906be. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Large","Routing Strategy":"GA"}. This record does not assert independent reproduction. ### BlockFFN PIQA Accuracy: 66.1. Result ID: 6a035d8b47f34ce924e906bd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large"}. This record does not assert independent reproduction. ### ReMoE PIQA Accuracy: 66.05. Result ID: 6a035d8b47f34ce924e906bb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large","Routing Strategy":"GA"}. This record does not assert independent reproduction. ### DeepSeek-V3 PIQA Accuracy: 65.94. Result ID: 6a035d8b47f34ce924e906ba. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large","Routing Strategy":"NG"}. This record does not assert independent reproduction. ### DECO PIQA Accuracy: 65.94. Result ID: 6a035d8b47f34ce924e906c0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Large","Routing Strategy":"NG"}. This record does not assert independent reproduction. ### TopP PIQA Accuracy: 64.53. Result ID: 6a035d8b47f34ce924e906b8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Large"}. This record does not assert independent reproduction. ### DECO PIQA Accuracy: 60.94. Result ID: 6a035d8b47f34ce924e906b1. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Small","MoE Normalization Variant":"NG"}. This record does not assert independent reproduction. ### BlockFFN PIQA Accuracy: 60.83. Result ID: 6a035d8b47f34ce924e906ae. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Small"}. This record does not assert independent reproduction. ### Dense PIQA Accuracy: 60.77. Result ID: 6a035d8b47f34ce924e906b0. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Small"}. This record does not assert independent reproduction. ### DeepSeek-V3 PIQA Accuracy: 60.66. Result ID: 6a035d8b47f34ce924e906ab. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Small","MoE Normalization Variant":"NG"}. This record does not assert independent reproduction. ### ReMoE PIQA Accuracy: 60.34. Result ID: 6a035d8b47f34ce924e906ac. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Small","MoE Normalization Variant":"GA"}. This record does not assert independent reproduction. ### TopP PIQA Accuracy: 60.17. Result ID: 6a035d8b47f34ce924e906a9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Small"}. This record does not assert independent reproduction. ### DeepSeek-V3 PIQA Accuracy: 60.12. Result ID: 6a035d8b47f34ce924e906aa. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Small","MoE Normalization Variant":"GA"}. This record does not assert independent reproduction. ### ReMoE PIQA Accuracy: 59.74. Result ID: 6a035d8b47f34ce924e906ad. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Scale":"Small","MoE Normalization Variant":"NG"}. This record does not assert independent reproduction. ### DECO PIQA Accuracy: 58.22. Result ID: 6a035d8b47f34ce924e906af. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Scale":"Small","MoE Normalization Variant":"GA"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.