Question about content of table2

microsoft / SoM

Set-of-Mark Prompting for GPT-4V and LMMs

MIT License

1.2k stars 96 forks source link

Question about content of table2 #15

Closed PeiMing1998 closed 11 months ago

PeiMing1998 commented 11 months ago

Impressive work! But I am a confused why in Table2 llava and shikera are deemed as no zero-shot ability models? Thank you.

jwyang commented 11 months ago

thanks, we consider them as non-zero-shot because they were trained with referring or grounding annotations, e.g., refcoco, flickr30, visual genome, etc.