| Paper | Code | pairwise accuracy | Accuracy (%) | ModelName | ReleaseDate |
|---|---|---|---|---|---|
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 75.7 | 54.4 | ViLBERT 12-in-1 | 2021-12-14 |
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 54.5 | GPT2 | 2021-12-14 | |
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 52.1 | CLIP | 2021-12-14 | |
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 49.5 | 50.0 | VisualBERT | 2021-12-14 |
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 47.2 | 50.0 | ViLBERT | 2021-12-14 |
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 46.8 | 49.8 | LXMERT | 2021-12-14 |
| VALSE: A Task-Independent Benchmark for Vision and Language Models Centered on Linguistic Phenomena | ✓ Link | 45.6 | GPT1 | 2021-12-14 |