Arithmetic on Basic Arithmetic
0.1921DC ScoreGPT-J
Evaluation Results
| Method | Links | ||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|
| GPT-JParameters=6B, Few-shot settings=5-shot2022.04 | 0.192 | 0.88 | 0.282 | 0.817 | 0.357 | 0.497 | 0.058 | 0.092 | 0.009 | 0.021 | |
| GPT-NeoXParameters=20B, Few-shot settings=5-shot2022.04 | 0.191 | 0.992 | 0.452 | 0.942 | 0.599 | 0.819 | 0.152 | 0.151 | 0.033 | 0.059 | |
| GPT-NeoXParameters=20B, Zero-shot=true2022.04 | 0.098 | 0.57 | 0.148 | 0.68 | 0.099 | 0.344 | 0.007 | 0.029 | 0 | 0.004 | |
| GPT-3 DaVinciZero-shot=true2022.04 | 0.098 | 0.769 | 0.198 | 0.58 | 0.342 | 0.483 | 0.04 | 0.075 | 0.006 | 0.008 | |
| GPT-JParameters=6B, Zero-shot=true2022.04 | 0.088 | 0.238 | 0.139 | 0.216 | 0.088 | 0.046 | 0.007 | 0.005 | 0.001 | 0 | |
| FairSeqModel Size=6.7B, Evaluation Protocol=5-shot2022.04 | 0.046 | 0.034 | 0.049 | 0.018 | 0.001 | 0.002 | 0 | 0 | 0 | 0 | |
| FairSeqModel Size=13B, Evaluation Protocol=5-shot2022.04 | 0.046 | 0.051 | 0.053 | 0.03 | 0.003 | 0.003 | 0 | 0 | 0 | 0 | |
| FairSeqModel Size=2.7B, Evaluation Protocol=5-shot2022.04 | 0.032 | 0.029 | 0.025 | 0.013 | 0.003 | 0.003 | 0 | 0.001 | 0 | 0 | |
| FairSeqModel Size=1.3B, Evaluation Protocol=5-shot2022.04 | 0.029 | 0.006 | 0.025 | 0.008 | 0.001 | 0.003 | 0 | 0 | 0 | 0 | |
| GPT-3 AdaZero-shot=true2022.04 | 0.029 | 0.006 | 0.022 | 0.013 | 0.001 | 0.001 | 0.001 | 0 | 0 | 0 | |
| FairSeqModel Size=355M, Evaluation Protocol=5-shot2022.04 | 0.024 | 0.004 | 0.025 | 0.011 | 0.002 | 0.004 | 0 | 0 | 0 | 0 | |
| GPT-3 CurieZero-shot=true2022.04 | 0.024 | 0.025 | 0.058 | 0.076 | 0.003 | 0.004 | 0.001 | 0 | 0 | 0 | |
| FairSeqModel Size=125M, Evaluation Protocol=5-shot2022.04 | 0.019 | 0.005 | 0.001 | 0.007 | 0.002 | 0.002 | 0 | 0.001 | 0 | 0 | |
| GPT-3 BabbageZero-shot=true2022.04 | 0.001 | 0.009 | 0.021 | 0.013 | 0.001 | 0.001 | 0 | 0 | 0 | 0 |