Predicting object attributes on PIGPEN-NLU (val)
81.8AccuracyPIGLET
Evaluation results
| Method | Links | |
|---|---|---|
| PIGLET2021.06 | 81.8 | |
| T5-11BApproach category=text-to-text2021.06 | 68.5 | |
| T5-3BApproach category=text-to-text2021.06 | 66.6 | |
| Alberti et al.2019Approach category=BERT-style, Dynamics pre-training=true2021.06 | 61.3 | |
| T5-LargeApproach category=text-to-text2021.06 | 56.5 | |
| T5-BaseApproach category=text-to-text2021.06 | 56 | |
| G&D2019Approach category=BERT-style, Dynamics pre-training=true2021.06 | 43.8 | |
| T5-SmallApproach category=text-to-text2021.06 | 39.9 | |
| No Change2021.06 | 27.4 | |
| GPT3-175Bfew-shot=true2021.06 | 23.8 | |
| G&D2019Approach category=BERT-style, Dynamics pre-training=false2021.06 | 15.1 | |
| Alberti et al.2019Approach category=BERT-style, Dynamics pre-training=false2021.06 | 9.7 |