# Physical Commonsense Reasoning on PIQA (Accuracy (%), Avg.) Source: [Physical Commonsense Reasoning on PIQA (Accuracy (%), Avg.)]() Collection index: [llms.txt]() Data updated: 2026-06-18T06:49:32.519Z Evaluation of model performance on the PIQA dataset for physical commonsense reasoning, reporting standard Accuracy percentage and an Average 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 | Physical Commonsense Reasoning | | Dataset (including any stated split) | PIQA | | Metric | Accuracy (%) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 36 of 36 | | 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 - [Accuracy (%)](): Higher is better; unit: PERCENT. - [Average Score](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (%) | Result date | Report paper | | --- | --- | --- | --- | | MoNE | 81.08 | Not recorded | 6a337b68206542251d7c4f03 | | He et al. | 81.08 | Not recorded | 6a337b68206542251d7c4f03 | | Baseline | 80.52 | Not recorded | 6a337b68206542251d7c4f03 | | Baseline | 80.2 | Not recorded | 6a337b68206542251d7c4f03 | | MoNE | 79.76 | Not recorded | 6a337b68206542251d7c4f03 | | Wanda | 79.65 | Not recorded | 6a337b68206542251d7c4f03 | | OursQ | 79.6 | Not recorded | 6a337b68206542251d7c4f03 | | EAC-MoE | 79.54 | Not recorded | 6a337b68206542251d7c4f03 | | OursQ | 79.54 | Not recorded | 6a337b68206542251d7c4f03 | | Wanda | 79.38 | Not recorded | 6a337b68206542251d7c4f03 | | PuzzleMoE | 78.7 | Not recorded | 6a337b68206542251d7c4f03 | | Wanda | 78.51 | Not recorded | 6a337b68206542251d7c4f03 | | He et al. | 78.5 | Not recorded | 6a337b68206542251d7c4f03 | | PuzzleMoE | 78.4 | Not recorded | 6a337b68206542251d7c4f03 | | Attribution-guided and Coverage-Maximized Expert-wise Pruning | 78.4 | Not recorded | 6a337b68206542251d7c4f03 | | He et al. | 78.2 | Not recorded | 6a337b68206542251d7c4f03 | | Attribution-guided and Coverage-Maximized Expert-wise Pruning | 77.97 | Not recorded | 6a337b68206542251d7c4f03 | | Wanda | 77.86 | Not recorded | 6a337b68206542251d7c4f03 | | C-Prune | 77.86 | Not recorded | 6a337b68206542251d7c4f03 | | EAC-MoE | 77.2 | Not recorded | 6a337b68206542251d7c4f03 | | FP | 75.79 | Not recorded | 6a264c8290966ade919be92f | | FP | 75.41 | Not recorded | 6a264c8290966ade919be92f | | MoNE | 75.3 | Not recorded | 6a337b68206542251d7c4f03 | | FAIR-Calib | 73.12 | Not recorded | 6a264c8290966ade919be92f | | FAIR-Calib | 73.07 | Not recorded | 6a264c8290966ade919be92f | | QuaRot | 72.52 | Not recorded | 6a264c8290966ade919be92f | | FlatQuant | 71.82 | Not recorded | 6a264c8290966ade919be92f | | FlatQuant | 71.65 | Not recorded | 6a264c8290966ade919be92f | | QuaRot | 71.11 | Not recorded | 6a264c8290966ade919be92f | | He et al. | 68.17 | Not recorded | 6a337b68206542251d7c4f03 | | Wanda | 64.15 | Not recorded | 6a337b68206542251d7c4f03 | | RTN | 63.33 | Not recorded | 6a264c8290966ade919be92f | | RTN | 61.7 | Not recorded | 6a264c8290966ade919be92f | | C-Prune | 58.92 | Not recorded | 6a337b68206542251d7c4f03 | | MoNE | 53.37 | Not recorded | 6a337b68206542251d7c4f03 | | MoNE | 49.51 | Not recorded | 6a337b68206542251d7c4f03 | ### MoNE Accuracy (%): 81.08. Result ID: 6a3394f4fbc373592c303cd4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"6.00GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### He et al. Accuracy (%): 81.08. Result ID: 6a3394f4fbc373592c303cd6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"6.00GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Baseline Accuracy (%): 80.52. Result ID: 6a3394f4fbc373592c303cc1. [Reporting paper]() Recorded evaluation configuration: {"Type":"None","Storage":"32.7GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Baseline Accuracy (%): 80.2. Result ID: 6a3394f4fbc373592c303cce. [Reporting paper]() Recorded evaluation configuration: {"Type":"None","Storage":"31.41GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### MoNE Accuracy (%): 79.76. Result ID: 6a3394f4fbc373592c303cd2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25%","Storage":"24.00GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Wanda Accuracy (%): 79.65. Result ID: 6a3394f4fbc373592c303ccf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25%","Storage":"31.41GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### OursQ Accuracy (%): 79.6. Result ID: 6a3394f4fbc373592c303ccd. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"6.26GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### EAC-MoE Accuracy (%): 79.54. Result ID: 6a3394f4fbc373592c303cc8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P11% Q3.03b","Storage":"7.19GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### OursQ Accuracy (%): 79.54. Result ID: 6a3394f4fbc373592c303cda. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"6.00GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Wanda Accuracy (%): 79.38. Result ID: 6a3394f4fbc373592c303cd1. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"7.70GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### PuzzleMoE Accuracy (%): 78.7. Result ID: 6a3394f4fbc373592c303cca. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"M25%","Storage":"25.04GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Wanda Accuracy (%): 78.51. Result ID: 6a3394f4fbc373592c303cc2. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25%","Storage":"32.7GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### He et al. Accuracy (%): 78.5. Result ID: 6a3394f4fbc373592c303cc7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"7.70GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### PuzzleMoE Accuracy (%): 78.4. Result ID: 6a3394f4fbc373592c303ccb. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"M50%","Storage":"17.34GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Attribution-guided and Coverage-Maximized Expert-wise Pruning Accuracy (%): 78.4. Result ID: 6a3394f4fbc373592c303cd9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"16.57GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### He et al. Accuracy (%): 78.2. Result ID: 6a3394f4fbc373592c303cd5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"16.57GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Attribution-guided and Coverage-Maximized Expert-wise Pruning Accuracy (%): 77.97. Result ID: 6a3394f4fbc373592c303ccc. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"17.34GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Wanda Accuracy (%): 77.86. Result ID: 6a3394f4fbc373592c303cc3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"32.7GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### C-Prune Accuracy (%): 77.86. Result ID: 6a3394f4fbc373592c303cd8. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25% Q4b","Storage":"6.00GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### EAC-MoE Accuracy (%): 77.2. Result ID: 6a3394f4fbc373592c303cc9. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P38% Q3.03b","Storage":"4.47GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### FP Accuracy (%): 75.79. Result ID: 6a265f94fbc373592c225eb1. [Reporting paper]() Recorded evaluation configuration: {"Model":"Dream-Instruct","Quantization":"Full Precision"}. This record does not assert independent reproduction. ### FP Accuracy (%): 75.41. Result ID: 6a265f94fbc373592c225eac. [Reporting paper]() Recorded evaluation configuration: {"Model":"Dream-Base","Quantization":"Full Precision"}. This record does not assert independent reproduction. ### MoNE Accuracy (%): 75.3. Result ID: 6a3394f4fbc373592c303cd3. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"16.57GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### FAIR-Calib Accuracy (%): 73.12. Result ID: 6a265f94fbc373592c225eb5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Dream-Instruct","Quantization":"W4A4"}. This record does not assert independent reproduction. ### FAIR-Calib Accuracy (%): 73.07. Result ID: 6a265f94fbc373592c225eb0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Dream-Base","Quantization":"W4A4"}. This record does not assert independent reproduction. ### QuaRot Accuracy (%): 72.52. Result ID: 6a265f94fbc373592c225eae. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Dream-Base","Quantization":"W4A4"}. This record does not assert independent reproduction. ### FlatQuant Accuracy (%): 71.82. Result ID: 6a265f94fbc373592c225eb4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Dream-Instruct","Quantization":"W4A4"}. This record does not assert independent reproduction. ### FlatQuant Accuracy (%): 71.65. Result ID: 6a265f94fbc373592c225eaf. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model":"Dream-Base","Quantization":"W4A4"}. This record does not assert independent reproduction. ### QuaRot Accuracy (%): 71.11. Result ID: 6a265f94fbc373592c225eb3. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model":"Dream-Instruct","Quantization":"W4A4"}. This record does not assert independent reproduction. ### He et al. Accuracy (%): 68.17. Result ID: 6a3394f4fbc373592c303cc6. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"17.34GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### Wanda Accuracy (%): 64.15. Result ID: 6a3394f4fbc373592c303cd0. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"31.41GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### RTN Accuracy (%): 63.33. Result ID: 6a265f94fbc373592c225eb2. [Reporting paper]() Recorded evaluation configuration: {"Model":"Dream-Instruct","Quantization":"W4A4"}. This record does not assert independent reproduction. ### RTN Accuracy (%): 61.7. Result ID: 6a265f94fbc373592c225ead. [Reporting paper]() Recorded evaluation configuration: {"Model":"Dream-Base","Quantization":"W4A4"}. This record does not assert independent reproduction. ### C-Prune Accuracy (%): 58.92. Result ID: 6a3394f4fbc373592c303cd7. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"16.57GB","Model Family":"DeepSeek-V2-Lite","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### MoNE Accuracy (%): 53.37. Result ID: 6a3394f4fbc373592c303cc4. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P25%","Storage":"25.04GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. ### MoNE Accuracy (%): 49.51. Result ID: 6a3394f4fbc373592c303cc5. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Type":"P50%","Storage":"17.34GB","Model Family":"Deepseek-MoE-16B","Evaluation Protocol":"Zero-shot"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.