{"id":6687,"date":"2024-03-30T08:47:24","date_gmt":"2024-03-30T00:47:24","guid":{"rendered":"https:\/\/www.1ai.net\/?p=6687"},"modified":"2024-03-30T08:47:24","modified_gmt":"2024-03-30T00:47:24","slug":"%e9%a9%ac%e6%96%af%e5%85%8b%e7%aa%81%e7%84%b6%e5%8f%91%e5%b8%83grok-1-5%ef%bc%81%e4%b8%8a%e4%b8%8b%e6%96%87%e9%95%bf%e5%ba%a6%e9%a3%99%e5%8d%8716%e5%80%8d%e5%92%8cgpt-4%e9%bd%90%e5%b9%b3","status":"publish","type":"post","link":"https:\/\/www.1ai.net\/en\/6687.html","title":{"rendered":"Musk suddenly released Grok 1.5! The context length soared 16 times and was on par with GPT-4"},"content":{"rendered":"<p>Elon<a href=\"https:\/\/www.1ai.net\/en\/tag\/%e9%a9%ac%e6%96%af%e5%85%8b\" title=\"[View articles tagged with [Musk]]\" target=\"_blank\" >Musk<\/a>Our AI startups<strong><a href=\"https:\/\/www.1ai.net\/en\/tag\/xai\" title=\"[View articles tagged with [xAI]]\" target=\"_blank\" >xA<\/a>Announcement of the official launch<a href=\"https:\/\/www.1ai.net\/en\/tag\/grok\" title=\"[See articles with [Grok] labels]\" target=\"_blank\" >Grok<\/a>-1.5,<\/strong>The official push didn\u2019t say anything, just threw out the link, with the main message being \u201cless is more\u201d.<\/p>\n<p><strong>What upgrades are there in Grok-1.5? There are two main aspects:<\/strong><\/p>\n<p><strong>1. Long context understanding<\/strong><\/p>\n<p>For the context window,<strong>Grok-1.5 directly increased it to 16 times the previous level.<\/strong>It increased from 8192 to 128k, which is on par with GPT-4.<\/p>\n<p>This means that Grok-1.5 can handle longer and more complex prompts while maintaining its ability to follow instructions.<\/p>\n<p>In the Needle in an Haystack (NIAH) evaluation, Grok-1.5 demonstrated powerful retrieval capabilities, retrieving embedded text in contexts up to 128K in length and achieving perfect retrieval results.<\/p>\n<p><strong>2. Ability and Reasoning<\/strong><\/p>\n<p>Grok-1.5<span class=\"spamTxt\">maximum<\/span>One of the improvements is the ability to handle programming and math-related tasks.<strong>It surpasses Grok-1, Mistral Large and Claude 2 in all aspects.<\/strong><\/p>\n<p>In mathematics, Grok-1.5 scored 50.6% on the MATH benchmark, surpassing the medium-sized Claude 3 Sonnet; and scored 90% on GSM8K.<\/p>\n<p><strong>In terms of programming, Grok-1.5 scored 74.1% on the HumanEval benchmark.<\/strong>It surpasses the medium-sized Claude 3 Sonnet, Gemini Pro1.5, and GPT-4, and is second only to the large-sized Claude 3 Opus.<\/p>\n<p class=\"article-content__img\"><img loading=\"lazy\" decoding=\"async\" class=\"alignnone size-full wp-image-6688\" title=\"2024032913380272810\" src=\"https:\/\/www.1ai.net\/wp-content\/uploads\/2024\/03\/2024032913380272810.jpg\" alt=\"2024032913380272810\" width=\"600\" height=\"253\" \/><\/p>","protected":false},"excerpt":{"rendered":"<p>Elon Musk's artificial intelligence startup xAI announced the official launch of Grok-1.5, the official push nothing to say, directly dump link, the main one word less things big \". Grok-1.5 what are the upgrades, mainly in two aspects: 1, long contextual understanding For the context window, Grok-1.5 directly improved to 16 times before, from 8192 to 128k growth, and GPT-4 flush. This also means that Grok-1.5 can handle longer and more complex cues while maintaining its ability to follow instructions. In the Needle in a Haystack (NIAH) evaluation, Grok-1.5 demonstrated powerful retrieval capabilities to retrieve embedded text in contexts up to 128K in length, obtaining perfect retrieval results. 2. Capabilities and<\/p>","protected":false},"author":1,"featured_media":0,"comment_status":"open","ping_status":"open","sticky":false,"template":"","format":"standard","meta":{"footnotes":""},"categories":[148,146],"tags":[364,356,355],"collection":[],"class_list":["post-6687","post","type-post","status-publish","format-standard","hentry","category-headline","category-news","tag-grok","tag-xai","tag-355"],"acf":[],"_links":{"self":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/6687","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts"}],"about":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/types\/post"}],"author":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/comments?post=6687"}],"version-history":[{"count":0,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/posts\/6687\/revisions"}],"wp:attachment":[{"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/media?parent=6687"}],"wp:term":[{"taxonomy":"category","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/categories?post=6687"},{"taxonomy":"post_tag","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/tags?post=6687"},{"taxonomy":"collection","embeddable":true,"href":"https:\/\/www.1ai.net\/en\/wp-json\/wp\/v2\/collection?post=6687"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}