VL-ICL Bench
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Vision-Language In-Context Learning | VL-ICL Bench Fast Open MiniImageNet (test) | Top-1 Accuracy100 | 14 | |
| Object-Property Induction | VL-ICL Bench OP_IND | Accuracy89.4 | 14 | |
| Few-shot Visual Question An Answering | VL-ICL Bench TextOCR (test) | Mean Accuracy26.4 | 13 | |
| Few-shot Visual Question Answering | VL-ICL Bench CLEVR (test) | Accuracy31.4 | 13 | |
| Operator Induction | VL-ICL Bench OP_IND (test) | Mean Accuracy47.7 | 13 | |
| Few-shot Visual Question Answering | VL-ICL Bench Open-MI (test) | Mean Accuracy77.9 | 13 |
Showing 6 of 6 rows