[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"tag-cross-attention":3},{"tag":4,"articles":9},{"id":5,"name":6,"slug":6,"article_count":7,"description_zh":8,"description_en":8},"19ec5cf0-5c97-48b0-8b6a-646937ddaa05","cross-attention",2,null,[10],{"id":11,"slug":12,"title":13,"summary":14,"category":15,"image_url":16,"cover_image":16,"language":17,"created_at":18},"5db60597-3ab6-4c21-90ec-13fa9bd30bd6","steerable-vi-vision-representations-en","Steerable ViT Features for Text-Guided Vision","A new vision representation lets text steer ViT features toward specific objects without giving up generic visual utility.","research","https:\u002F\u002Fxxdpdyhzhpamafnrdkyq.supabase.co\u002Fstorage\u002Fv1\u002Fobject\u002Fpublic\u002Fcovers\u002Finline-1775196409641-4e2e.png","en","2026-04-03T06:06:34.954101+00:00"]