{"id":6758,"date":"2025-05-19T06:41:00","date_gmt":"2025-05-18T21:41:00","guid":{"rendered":"https:\/\/devneko.jp\/wordpress\/?p=6758"},"modified":"2025-05-17T14:58:13","modified_gmt":"2025-05-17T05:58:13","slug":"seed1-5-vl-qwen3-mimo-minimax-speech-aya-vision-blip3-o","status":"publish","type":"post","link":"https:\/\/devneko.jp\/wordpress\/?p=6758","title":{"rendered":"Seed1.5-VL, Qwen3, MiMo,  MiniMax-Speech, Aya Vision, BLIP3-o"},"content":{"rendered":"\n<p class=\"wp-block-paragraph\">Bytedance\u306eSeek 1.5 VL\u3001Alibaba\u306eQwen3, Xiaomi\u306eMiMo\u3001MiniMax\u306eMiniMaz-Speech\u3068\u5148\u9031\u306f\u4e2d\u56fd\u306e\u7814\u7a76\u6a5f\u95a2\u304b\u3089\u306e\u8ad6\u6587\u516c\u958b\u304c\u591a\u304b\u3063\u305f\u3002\u307e\u305f\u3001Cohere\u306eAya Vision\u3001Salesforce\u306eBLIP3-o\u8ad6\u6587\u306e\u516c\u958b\u3082\u3042\u308a\u3001LLM\u3001MLLM\u95a2\u9023\u306fOpenAI\u4e00\u5f37\u3068\u3044\u3046\u72b6\u614b\u3067\u306f\u306a\u304f\u306a\u3063\u3066\u3044\u308b\u3002\u8457\u8005\u30ea\u30b9\u30c8\u3092\u898b\u308b\u3068\u6709\u529b\u306a\u7814\u7a76\u8005\u304c\u8907\u6570\u6240\u5c5e\u3059\u308b\u5927\u898f\u6a21\u306a\u30c1\u30fc\u30e0\u3067\u30e2\u30c7\u30eb\u69cb\u7bc9\u3092\u884c\u3063\u3066\u3044\u308b\u3088\u3046\u306b\u898b\u3048\u308b\u3002<\/p>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Seed1.5-VL Technical Report\u00a0<\/strong>[237.8]<br>Seed1.5-VL\u306f\u3001\u6c4e\u7528\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u7406\u89e3\u3068\u63a8\u8ad6\u3092\u4fc3\u9032\u3059\u308b\u305f\u3081\u306b\u8a2d\u8a08\u3055\u308c\u305f\u30d3\u30b8\u30e7\u30f3\u8a00\u8a9e\u57fa\u76e4\u30e2\u30c7\u30eb\u3067\u3042\u308b\u3002 \u5e45\u5e83\u3044\u30d1\u30d6\u30ea\u30c3\u30af\u306aVLM\u30d9\u30f3\u30c1\u30de\u30fc\u30af\u3068\u5185\u90e8\u8a55\u4fa1\u30b9\u30a4\u30fc\u30c8\u3067\u5f37\u529b\u306a\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u3092\u63d0\u4f9b\u3059\u308b\u3002 GUI\u5236\u5fa1\u3084\u30b2\u30fc\u30e0\u30d7\u30ec\u30a4\u3068\u3044\u3063\u305f\u30a8\u30fc\u30b8\u30a7\u30f3\u30c8\u4e2d\u5fc3\u306e\u30bf\u30b9\u30af\u3067\u306f\u3001Seed1.5-VL\u306fOpenAI CUA\u3084Claude 3.7\u306a\u3069\u3001\u4e3b\u8981\u306a\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u30b7\u30b9\u30c6\u30e0\u3088\u308a\u512a\u308c\u3066\u3044\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2505.07062v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2505.07062v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Sun, 11 May 2025 17:28:30 GMT)<\/li>\n\n\n\n<li>\u300cDespite its relatively compact architecture, it delivers strong performance across a wide spectrum of public VLM benchmarks and internal evaluation suites, achieving the state-of-the-art performance on 38 out of 60 public benchmarks. Moreover, in agent-centric tasks such as GUI control and gameplay, Seed1.5-VL outperforms leading multimodal systems, including OpenAI CUA and Claude 3.7.\u300d\u3092\u4e3b\u5f35\u3059\u308bMLLM<\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Qwen3 Technical Report\u00a0<\/strong>[138.0]<br>Qwen\u30e2\u30c7\u30eb\u30d5\u30a1\u30df\u30ea\u306e\u6700\u65b0\u30d0\u30fc\u30b8\u30e7\u30f3\u3067\u3042\u308bQwen3\u3092\u7d39\u4ecb\u3057\u307e\u3059\u3002 Qwen3\u306f\u3001\u6027\u80fd\u3001\u52b9\u7387\u3001\u591a\u8a00\u8a9e\u6a5f\u80fd\u3092\u5411\u4e0a\u3055\u305b\u308b\u305f\u3081\u306b\u8a2d\u8a08\u3055\u308c\u305f\u4e00\u9023\u306e\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb(LLM)\u304b\u3089\u69cb\u6210\u3055\u308c\u3066\u3044\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2505.09388v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2505.09388v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Wed, 14 May 2025 13:41:34 GMT)<\/li>\n\n\n\n<li>Qwen\uff08<a href=\"https:\/\/devneko.jp\/wordpress\/?p=6660\">Qwen3, Phi-4 reasoning, MiMo 7B, OLMo2 1B, Mellum 4B \u2013 arXiv\u6700\u65b0\u8ad6\u6587\u306e\u7d39\u4ecb<\/a>\uff09\u306b\u95a2\u3057\u3066arXiv\u306b\u6295\u7a3f\u3055\u308c\u305f\u8ad6\u6587<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/QwenLM\/Qwen3\">GitHub &#8211; QwenLM\/Qwen3: Qwen3 is the large language model series developed by Qwen team, Alibaba Cloud.<\/a><\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>MiMo: Unlocking the Reasoning Potential of Language Model &#8212; From Pretraining to Posttraining\u00a0<\/strong>[66.1]<br>\u63d0\u6848\u3059\u308bMiMo-7B\u306f,\u5b66\u7fd2\u524d\u306e\u6bb5\u968e\u3068\u5b66\u7fd2\u5f8c\u306e\u6bb5\u968e\u306b\u307e\u305f\u304c\u3063\u3066\u6700\u9069\u5316\u3055\u308c\u305f,\u63a8\u8ad6\u30bf\u30b9\u30af\u306e\u305f\u3081\u306e\u5927\u898f\u6a21\u8a00\u8a9e\u30e2\u30c7\u30eb\u3067\u3042\u308b\u3002 MiMo-7B-Base\u306f25\u5146\u306e\u30c8\u30fc\u30af\u30f3\u3067\u4e8b\u524d\u8a13\u7df4\u3055\u308c\u3066\u304a\u308a\u3001\u6027\u80fd\u306e\u5411\u4e0a\u3068\u63a8\u8ad6\u901f\u5ea6\u306e\u9ad8\u901f\u5316\u3092\u76ee\u6a19\u3068\u3057\u3066\u3044\u308b\u3002 \u6700\u5f8c\u306eRL\u30c1\u30e5\u30fc\u30cb\u30f3\u30b0\u30e2\u30c7\u30eb\u3067\u3042\u308bMiMo-7B-RL\u306f\u3001OpenAI o1-mini\u306e\u6027\u80fd\u3092\u4e0a\u56de\u308a\u3001\u6570\u5b66\u3001\u30b3\u30fc\u30c9\u3001\u4e00\u822c\u7684\u306a\u63a8\u8ad6\u30bf\u30b9\u30af\u306b\u304a\u3044\u3066\u512a\u308c\u305f\u30d1\u30d5\u30a9\u30fc\u30de\u30f3\u30b9\u3092\u9054\u6210\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2505.07608v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2505.07608v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Mon, 12 May 2025 14:30:11 GMT)<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/XiaomiMiMo\/MiMo\">GitHub &#8211; XiaomiMiMo\/MiMo: MiMo: Unlocking the Reasoning Potential of Language Model \u2013 From Pretraining to Posttraining<\/a><\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>Aya Vision: Advancing the Frontier of Multilingual Multimodality\u00a0<\/strong>[16.0]<br>\u9ad8\u54c1\u8cea\u3067\u591a\u69d8\u306a\u591a\u8a00\u8a9e\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u547d\u4ee4\u30c7\u30fc\u30bf\u3092\u30ad\u30e5\u30ec\u30fc\u30c8\u3059\u308b\u5408\u6210\u30a2\u30ce\u30c6\u30fc\u30b7\u30e7\u30f3\u30d5\u30ec\u30fc\u30e0\u30ef\u30fc\u30af\u3092\u958b\u767a\u3057\u305f\u3002 \u307e\u305f,\u7834\u6ec5\u7684\u5fd8\u308c\u3092\u7de9\u548c\u3059\u308b\u30af\u30ed\u30b9\u30e2\u30fc\u30c0\u30eb\u30e2\u30c7\u30eb\u30de\u30fc\u30b8\u624b\u6cd5\u3092\u63d0\u6848\u3059\u308b\u3002 \u6211\u3005\u306e\u7814\u7a76\u306f\u3001\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u30d5\u30ed\u30f3\u30c6\u30a3\u30a2\u306b\u304a\u3051\u308b\u591a\u8a00\u8a9e\u7684\u306a\u9032\u6b69\u3092\u524d\u9032\u3055\u305b\u3001\u8a08\u7b97\u306e\u5fc5\u8981\u6027\u3092\u52b9\u679c\u7684\u306b\u66f2\u3052\u308b\u6280\u8853\u306b\u95a2\u3059\u308b\u6d1e\u5bdf\u3092\u63d0\u4f9b\u3059\u308b\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2505.08751v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2505.08751v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Tue, 13 May 2025 17:03:48 GMT)<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/huggingface.co\/collections\/CohereLabs\/cohere-labs-aya-vision-67c4ccd395ca064308ee1484\">Cohere Labs Aya Vision &#8211; a CohereLabs Collection<\/a><\/li>\n<\/ul>\n\n\n\n<ul class=\"wp-block-list\">\n<li><strong>BLIP3-o: A Family of Fully Open Unified Multimodal Models-Architecture, Training and Dataset\u00a0<\/strong>[140.2]<br>\u672c\u7a3f\u3067\u306f,\u62e1\u6563\u5909\u63db\u5668\u3092\u7528\u3044\u3066\u610f\u5473\u7684\u306b\u30ea\u30c3\u30c1\u306aCLIP\u753b\u50cf\u7279\u5fb4\u3092\u751f\u6210\u3059\u308b\u624b\u6cd5\u3092\u63d0\u6848\u3059\u308b\u3002 \u753b\u50cf\u7406\u89e3\u306e\u305f\u3081\u306e\u7d71\u5408\u30e2\u30c7\u30eb\u30d5\u30a1\u30fc\u30b9\u30c8\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u3068\u753b\u50cf\u751f\u6210\u306e\u305f\u3081\u306e\u9010\u6b21\u4e8b\u524d\u5b66\u7fd2\u6226\u7565\u306f\u3001\u5b9f\u7528\u7684\u306a\u5229\u70b9\u3092\u3082\u305f\u3089\u3059\u3002 \u9769\u65b0\u7684\u306a\u30e2\u30c7\u30eb\u8a2d\u8a08\u3001\u30c8\u30ec\u30fc\u30cb\u30f3\u30b0\u30ec\u30b7\u30d4\u3001\u30c7\u30fc\u30bf\u30bb\u30c3\u30c8\u306b\u57fa\u3065\u3044\u3066\u3001\u6700\u5148\u7aef\u306e\u7d71\u4e00\u30de\u30eb\u30c1\u30e2\u30fc\u30c0\u30eb\u30e2\u30c7\u30eb\u306e\u30b9\u30a4\u30fc\u30c8\u3067\u3042\u308bBLIP3-o\u3092\u958b\u767a\u3057\u307e\u3059\u3002<br><a href=\"http:\/\/arxiv.org\/abs\/2505.09568v1\">\u8ad6\u6587<\/a>\u00a0\u00a0<a href=\"https:\/\/fugumt.com\/fugumt\/paper_check\/2505.09568v1\">\u53c2\u8003\u8a33\uff08\u30e1\u30bf\u30c7\u30fc\u30bf\uff09<\/a>\u00a0 \u00a0(Wed, 14 May 2025 17:11:07 GMT)<\/li>\n\n\n\n<li>\u30ea\u30dd\u30b8\u30c8\u30ea\u306f<a href=\"https:\/\/github.com\/JiuhaiChen\/BLIP3o\">GitHub &#8211; JiuhaiChen\/BLIP3o<\/a>\u3001<a href=\"https:\/\/huggingface.co\/BLIP3o\/BLIP3o-Model\">BLIP3o\/BLIP3o-Model \u00b7 Hugging Face<\/a><\/li>\n<\/ul>\n","protected":false},"excerpt":{"rendered":"<p>Bytedance\u306eSeek 1.5 VL\u3001Alibaba\u306eQwen3, Xiaomi\u306eMiMo\u3001MiniMax\u306eMiniMaz-Speech\u3068\u5148\u9031\u306f\u4e2d\u56fd\u306e\u7814\u7a76\u6a5f\u95a2\u304b\u3089\u306e\u8ad6\u6587\u516c\u958b\u304c\u591a\u304b\u3063\u305f\u3002\u307e\u305f\u3001Cohere\u306eAya V &hellip; <a href=\"https:\/\/devneko.jp\/wordpress\/?p=6758\" class=\"more-link\"><span class=\"screen-reader-text\">&#8220;Seed1.5-VL, Qwen3, MiMo,  MiniMax-Speech, Aya Vision, BLIP3-o&#8221; \u306e<\/span>\u7d9a\u304d\u3092\u8aad\u3080<\/a><\/p>\n","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[2],"tags":[223,251],"class_list":["post-6758","post","type-post","status-publish","format-standard","hentry","category-arxiv","tag-llm","tag-mllm"],"_links":{"self":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6758","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcomments&post=6758"}],"version-history":[{"count":1,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6758\/revisions"}],"predecessor-version":[{"id":6759,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=\/wp\/v2\/posts\/6758\/revisions\/6759"}],"wp:attachment":[{"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fmedia&parent=6758"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Fcategories&post=6758"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/devneko.jp\/wordpress\/index.php?rest_route=%2Fwp%2Fv2%2Ftags&post=6758"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}