Manli Shu

Latest

• Test-time prompt tuning for zero-shot generalization in vision-language models