Sentence Simplification on TurkCorpus English (test)
42.62SARIMUSS
Evaluation results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| MUSSTraining Paradigm=Supervised2020.05 | 42.62 | 6.98 | — | — | — | — | |
| MUSSTraining Paradigm=Supervised, Data augmentation=+ mined data2020.05 | 42.53 | 7.6 | — | — | — | — | |
| Martin et al.variant=2020b2021.03 | 42.53 | 7.6 | — | — | — | — | |
| TST-FINALmodel_version=final2021.03 | 41.46 | 7.87 | 6.96 | 47.87 | 69.56 | — | |
| ACCESSTraining Paradigm=Supervised2020.05 | 41.38 | 7.29 | — | — | — | — | |
| Martin et al.variant=2020a2021.03 | 41.38 | 7.29 | — | — | — | — | |
| MUSSTraining Paradigm=Unsupervised2020.05 | 40.85 | 8.79 | — | — | — | — | |
| Zhao et al.variant=2018a2021.03 | 40.42 | 7.79 | 5.72 | 42.23 | 73.41 | — | |
| ControlTS2023.05 | 40.4 | — | 6.2 | 44.5 | 70.4 | 95.9 | |
| Gold Reference2020.05 | 40.04 | 8.77 | — | — | — | — | |
| Reference Baseline2021.03 | 40.02 | 8.77 | 6.21 | 70.15 | 43.69 | — | |
| Xu et al.2021.03 | 39.96 | 7.29 | 5.96 | 41.42 | 72.52 | — | |
| DMASS-DCSSTraining Paradigm=Supervised2020.05 | 39.92 | 7.73 | — | — | — | — | |
| SimpleBART2023.05 | 39.5 | — | 6.6 | 47.2 | 64.6 | 97.2 | |
| SimpleBART2023.05 | 39.5 | — | 6.6 | 47.2 | 64.6 | 97.2 | |
| TST-BASEmodel_version=base2021.03 | 39.17 | 8.08 | 3.62 | 41.61 | 72.29 | — | |
| BART-CPtraining_strategy=continued pre-training with random masking2023.05 | 38.6 | — | 5.4 | 45.9 | 64.6 | 96.7 | |
| BART2023.05 | 38.3 | — | 5.6 | 44 | 65.4 | 97.3 | |
| Dong et al.2021.03 | 38.22 | 7.3 | 3.36 | 39.15 | 72.13 | — | |
| Mallinson et al.2021.03 | 38.13 | 8.98 | 3.55 | 40.45 | 70.39 | — | |
| EditNTS2023.05 | 37.9 | — | 3.4 | 43.1 | 67.3 | 95 | |
| T5Model size=base2023.05 | 37.8 | — | 4.2 | 35.6 | 73.5 | 98.2 | |
| EditNTSTraining Paradigm=Supervised2020.05 | 37.66 | 8.38 | — | — | — | — | |
| Guo et al.2021.03 | 37.45 | 7.41 | — | — | — | — | |
| RM+EX+LS+ROTraining Paradigm=Unsupervised2020.05 | 37.27 | 7.33 | — | — | — | — | |
| Zhang and Lapata2021.03 | 37.27 | 6.62 | — | — | — | — | |
| Zhao et al.variant=2020b2021.03 | 37.25 | — | 2.87 | 40.06 | 68.82 | — | |
| Qiang2021.03 | 37.21 | 6.56 | — | — | — | — | |
| Alva-Manchego et al.2021.03 | 37.08 | 5.35 | 2.94 | 43.2 | 65.1 | — | |
| Dress-LSTraining Paradigm=Supervised2020.05 | 36.97 | 7.66 | — | — | — | — | |
| Vu et al.2021.03 | 36.88 | — | — | — | — | — | |
| FLAN-T5variant=base, Evaluation Protocol=Fine-tuned2023.05 | 36.5 | — | 3.8 | 31.3 | 74.4 | 90.1 | |
| UNTSTraining Paradigm=Unsupervised2020.05 | 36.29 | 7.6 | — | — | — | — | |
| Nisioi et al.2021.03 | 35.66 | 8.42 | 2.99 | 28.96 | 75.02 | — | |
| Surya et al.2021.03 | 34.96 | — | — | — | — | — | |
| Truncate Baseline2020.05 | 33.1 | 7.91 | — | — | — | — | |
| BTRLTSTraining Paradigm=Unsupervised2020.05 | 33.09 | 8.39 | — | — | — | — | |
| GPT-3.5variant=Turbo-0301, Evaluation Protocol=Zero-shot2023.05 | 32.4 | — | 10.4 | 43.4 | 43.4 | 89.6 | |
| FLAN-T5variant=XL, Evaluation Protocol=Zero-shot2023.05 | 31.5 | — | 1 | 29.6 | 64.1 | 89.2 | |
| LLAMAvariant=7B, Evaluation Protocol=Zero-shot2023.05 | 29.3 | — | 2.3 | 16.3 | 69.3 | 87.3 | |
| Identity Baseline2020.05 | 26.29 | 10.02 | — | — | — | — |