@mia_chiquier : Multimodal pre-trained models, such as CLIP, are popular for zero-shot classification due to their open-vocabulary flexibility and high performance, but how would you classify images that don’t have obvious names using CLIP? • TwiCopy

Mia Chiquier

@mia_chiquier

+ Follow

Machine Learning PhD student at @Columbia. Focused on methods that help us learn from models i.e. discovery and interpretability.

ID: 1241365781595787269

linkhttp://www.cs.columbia.edu/~mia.chiquier/ calendar_today21-03-2020 14:07:26

39 Tweet

347 Followers

331 Following

Mia Chiquier

@mia_chiquier

5 months ago

Multimodal pre-trained models, such as CLIP, are popular for zero-shot classification due to their open-vocabulary flexibility and high performance, but how would you classify images that don’t have obvious names using CLIP?