# Physical Common Sense Reasoning on PIQA 2020 Source: [Physical Common Sense Reasoning on PIQA 2020]() Collection index: [llms.txt]() Data updated: 2026-06-08T07:32:24.034Z Evaluation of model performance on the PIQA 2020 dataset for physical common sense reasoning, reporting normalized accuracy (Acc\_norm). 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 Common Sense Reasoning | | Dataset (including any stated split) | PIQA 2020 | | Metric | Accuracy (Normalized) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 26 of 26 | | 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 (Normalized)](): Higher is better; unit: PERCENT. ## Recorded results | Method | Accuracy (Normalized) | Result date | Report paper | | --- | --- | --- | --- | | Baseline | 80.96 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale | 79.54 | Not recorded | 6a264c6490966ade919be87b | | ASVD+ | 79.49 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale (KD) | 79.38 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale | 77.75 | Not recorded | 6a264c6490966ade919be87b | | Baseline | 77.69 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale (KD) | 77.64 | Not recorded | 6a264c6490966ade919be87b | | ASVD+ | 76.01 | Not recorded | 6a264c6490966ade919be87b | | SVD-LLM | 75.57 | Not recorded | 6a264c6490966ade919be87b | | SVD-LLM | 75.24 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale | 73.94 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale (KD) | 73.88 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale | 73.78 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale (KD) | 73.72 | Not recorded | 6a264c6490966ade919be87b | | SVD-LLM | 72.69 | Not recorded | 6a264c6490966ade919be87b | | ASVD+ | 71.98 | Not recorded | 6a264c6490966ade919be87b | | SVD-LLM | 70.67 | Not recorded | 6a264c6490966ade919be87b | | ASVD+ | 69.53 | Not recorded | 6a264c6490966ade919be87b | | SVD-LLM | 66.76 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale | 65.07 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale (KD) | 64.69 | Not recorded | 6a264c6490966ade919be87b | | SVD-LLM | 62.13 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale (KD) | 61.81 | Not recorded | 6a264c6490966ade919be87b | | SigmaScale | 61.59 | Not recorded | 6a264c6490966ade919be87b | | ASVD+ | 59.3 | Not recorded | 6a264c6490966ade919be87b | | ASVD+ | 57.89 | Not recorded | 6a264c6490966ade919be87b | ### Baseline Accuracy (Normalized): 80.96. Result ID: 6a266f37fbc373592c22d54d. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"1.00x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale Accuracy (Normalized): 79.54. Result ID: 6a266f37fbc373592c22d550. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### ASVD+ Accuracy (Normalized): 79.49. Result ID: 6a266f37fbc373592c22d54f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale (KD) Accuracy (Normalized): 79.38. Result ID: 6a266f37fbc373592c22d551. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale Accuracy (Normalized): 77.75. Result ID: 6a266f37fbc373592c22d55d. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### Baseline Accuracy (Normalized): 77.69. Result ID: 6a266f37fbc373592c22d55a. [Reporting paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"1.00x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale (KD) Accuracy (Normalized): 77.64. Result ID: 6a266f37fbc373592c22d55e. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### ASVD+ Accuracy (Normalized): 76.01. Result ID: 6a266f37fbc373592c22d55c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SVD-LLM Accuracy (Normalized): 75.57. Result ID: 6a266f37fbc373592c22d55b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SVD-LLM Accuracy (Normalized): 75.24. Result ID: 6a266f37fbc373592c22d54e. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.90x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale Accuracy (Normalized): 73.94. Result ID: 6a266f37fbc373592c22d561. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale (KD) Accuracy (Normalized): 73.88. Result ID: 6a266f37fbc373592c22d562. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale Accuracy (Normalized): 73.78. Result ID: 6a266f37fbc373592c22d554. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale (KD) Accuracy (Normalized): 73.72. Result ID: 6a266f37fbc373592c22d555. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SVD-LLM Accuracy (Normalized): 72.69. Result ID: 6a266f37fbc373592c22d55f. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### ASVD+ Accuracy (Normalized): 71.98. Result ID: 6a266f37fbc373592c22d560. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SVD-LLM Accuracy (Normalized): 70.67. Result ID: 6a266f37fbc373592c22d552. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### ASVD+ Accuracy (Normalized): 69.53. Result ID: 6a266f37fbc373592c22d553. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.75x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SVD-LLM Accuracy (Normalized): 66.76. Result ID: 6a266f37fbc373592c22d563. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale Accuracy (Normalized): 65.07. Result ID: 6a266f37fbc373592c22d565. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale (KD) Accuracy (Normalized): 64.69. Result ID: 6a266f37fbc373592c22d566. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SVD-LLM Accuracy (Normalized): 62.13. Result ID: 6a266f37fbc373592c22d556. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale (KD) Accuracy (Normalized): 61.81. Result ID: 6a266f37fbc373592c22d559. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### SigmaScale Accuracy (Normalized): 61.59. Result ID: 6a266f37fbc373592c22d558. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### ASVD+ Accuracy (Normalized): 59.3. Result ID: 6a266f37fbc373592c22d564. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen3-8B","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. ### ASVD+ Accuracy (Normalized): 57.89. Result ID: 6a266f37fbc373592c22d557. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Llama 3.1 8B Instruct","Compression Ratio":"0.50x","Zero-shot":"true"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.