
{"id":13508,"date":"2026-10-17T02:43:58","date_gmt":"2026-10-16T18:43:58","guid":{"rendered":"https:\/\/infernews.com\/blog\/worldcast\/"},"modified":"2026-10-17T02:43:58","modified_gmt":"2026-10-16T18:43:58","slug":"worldcast","status":"publish","type":"post","link":"https:\/\/infernews.com\/blog\/worldcast\/","title":{"rendered":"WorldCast \u9810\u6e2c\u4e26\u751f\u6210\u4e16\u754c\u72c0\u614b"},"content":{"rendered":"\n<figure class=\"wp-block-image size-large\"><img decoding=\"async\" src=\"https:\/\/infernews.com\/blog\/wp-content\/uploads\/2026\/10\/WorldCast.jpg\" alt=\"(a) The four-level capability tree of visual generation. (b) The WorldCast benchmark and overall results.\"><\/figure>\n\n\n\n<p class=\"wp-block-paragraph\">\u7576\u4efb\u52d9\u8981\u6c42\u6a21\u578b\u6839\u64da\u89c0\u5bdf\u3001\u7269\u4ef6\u53c3\u7167\u548c\u64cd\u4f5c\u793a\u7bc4\uff0c\u9810\u6e2c\u4e00\u500b\u904e\u7a0b\u63a5\u4e0b\u4f86\u6703\u600e\u6a23\u767c\u5c55\uff0c\u55ae\u5f35\u5716\u7247\u7684\u5408\u7406\u6027\u5df2\u7d93\u4e0d\u5920\u3002WorldCast \u5c6c\u65bc\u8996\u89ba\u751f\u6210\u57fa\u6e96\u6e2c\u8a66\uff0c\u5be6\u969b\u8655\u7406\u7684\u662f\u591a\u6a21\u614b\u6a21\u578b\u5982\u4f55\u7406\u89e3\u4e0d\u540c\u53c3\u7167\u5728\u4f55\u6642\u751f\u6548\uff0c\u4e26\u751f\u6210\u4e00\u9023\u4e32\u6709\u56e0\u679c\u95dc\u4fc2\u7684\u4e16\u754c\u72c0\u614b\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u5b83\u628a\u80fd\u529b\u5206\u6210 L1 \u81f3 L4\uff1a\u7531\u975c\u614b text-to-image\uff0c\u63a8\u9032\u81f3\u7de8\u8f2f\u3001\u4e3b\u9ad4\u53c3\u7167\u548c text-to-video\uff0c\u518d\u9032\u5165\u8907\u96dc\u9650\u5236\u3001\u7269\u7406\u5408\u7406\u6027\u53ca\u628a\u73fe\u5be6\u60c5\u5883\u8f49\u5316\u70ba\u884c\u52d5\u3002WorldCast \u805a\u7126\u8f03\u5c11\u88ab\u91cf\u5ea6\u7684 L3\u3001L4\uff0c\u6db5\u84cb 320 \u500b\u4eba\u5de5\u8a2d\u8a08\u6848\u4f8b\u30011,352 \u500b\u8996\u89ba\u72c0\u614b\u53ca 8 \u985e\u4efb\u52d9\uff0c\u4e26\u540c\u6642\u6e2c\u8a66 unified systems\u3001multi-reference image models \u548c video models\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u2022 \u8a55\u5206\u628a grid\u3001interleaved \u548c video \u8f38\u51fa\u8996\u70ba\u6709\u6b21\u5e8f\u7684\u72c0\u614b\u5e8f\u5217<br \/>\n\u2022 \u53c3\u7167\u5167\u5bb9\u53ef\u4ee5\u662f\u89c0\u5bdf\u72c0\u614b\u3001\u76ee\u6a19\u3001\u898f\u5247\u3001\u7269\u6599\u7279\u6027\u6216\u793a\u7bc4\u7a0b\u5e8f<br \/>\n\u2022 12 \u500b\u4ee3\u8868\u6027\u6a21\u578b\u63a5\u53d7\u540c\u4e00\u5957\u8a55\u4f30\u5354\u8b70\u53ca Judge rubrics<br \/>\n\u2022 GPT-Image-2.5 System \u53d6\u5f97 72.8 \u5206\uff0c\u53cd\u6620\u524d\u6cbf\u6a21\u578b\u5df2\u51fa\u73fe\u65e9\u671f\u4e16\u754c\u72c0\u614b\u9810\u6e2c\u80fd\u529b<\/p>\n\n\n\n<p class=\"wp-block-paragraph\">\u53ef\u5f9e Hugging Face Dataset \u53d6\u5f97\u8cc7\u6599\u6e2c\u8a66\uff0c\u914d\u5408\u9805\u76ee\u63d0\u4f9b\u7684 evaluation toolkit \u548c Evaluation Guide \u91cd\u73fe\u8a55\u5206\uff1b\u6a21\u578b\u4ea6\u53ef\u8f38\u51fa\u9010\u683c\u5716\u50cf\u3001\u4ea4\u932f\u5716\u50cf\u6216\u5f71\u7247\u3002\u7d50\u679c\u5dee\u8ddd\u6703\u5728\u8f03\u9ad8\u5c64\u6b21\u4efb\u52d9\u64f4\u5927\uff0cinterleaved generation \u5c0d\u5716\u50cf\u6a21\u578b\u6709\u5e6b\u52a9\uff0c\u4f46\u8a55\u6e2c\u4ecd\u96c6\u4e2d\u65bc\u4eba\u5de5\u8a2d\u8a08\u6848\u4f8b\uff0c\u672a\u7b49\u540c\u65bc\u771f\u5be6\u74b0\u5883\u4e2d\u7684\u9577\u671f\u898f\u5283\u80fd\u529b\u3002<\/p>\n\n\n\n<p class=\"wp-block-paragraph\"><a href=\"https:\/\/teal024.github.io\/WorldCast\/\" rel=\"noopener noreferrer\"><strong>\u9805\u76ee\u4e3b\u9801<\/strong><\/a> \u00b7 <a href=\"https:\/\/github.com\/teal024\/WorldCast\" rel=\"noopener noreferrer\"><strong>GitHub<\/strong><\/a><\/p>\n","protected":false},"excerpt":{"rendered":"<p>WorldCast \u628a\u8996\u89ba\u751f\u6210\u62c9\u5230\u591a\u6b65\u63a8\u6f14\uff1a\u6a21\u578b\u8981\u4ea4\u4ee3\u4e16\u754c\u5982\u4f55\u9010\u6b65\u8b8a\u5316\uff0c\u800c\u4e0d\u53ea\u662f\u4ea4\u51fa\u4e00\u5f35\u5408\u7406\u7d42\u5c40\u5716\u3002<\/p>\n","protected":false},"author":8,"featured_media":13507,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"wpai_generated_summary":"","wpai_meta_description":"","footnotes":""},"categories":[133,164,162,76,119,141,31,30,129,199],"tags":[],"class_list":["post-13508","post","type-post","status-publish","format-standard","hentry","category-133","category-164","category-ai-productions","category-76","category-119","category-141","category-video","category-image","category-txt2img","category-dataset-"],"_links":{"self":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/13508","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/users\/8"}],"replies":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/comments?post=13508"}],"version-history":[{"count":0,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/posts\/13508\/revisions"}],"wp:featuredmedia":[{"embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media\/13507"}],"wp:attachment":[{"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/media?parent=13508"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/categories?post=13508"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/infernews.com\/blog\/wp-json\/wp\/v2\/tags?post=13508"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}