Integrated Structural Prompt Learning for Vision-Language Models
Open paper
Benchmark
Model
Rank
Results
prompt-engineering-on-imagenet-a
ISP
–
Top-1 accuracy %: 49.90
prompt-engineering-on-imagenet-r
ISP
–
Top-1 accuracy %: 77.40
prompt-engineering-on-imagenet-v2
ISP
–
Top-1 accuracy %: 64.50
Rank counts only results with a code link.