TE
科技回声
首页24小时热榜最新最佳问答展示工作
GitHubTwitter
首页

科技回声

基于 Next.js 构建的科技新闻平台,提供全球科技新闻和讨论内容。

GitHubTwitter

首页

首页最新最佳问答展示工作

资源链接

HackerNews API原版 HackerNewsNext.js

© 2025 科技回声. 版权所有。

Sparks of AGI? – Analyzing GPT-4 and the Latest GPT/LLM Models [sentdex]

2 点作者 pabo大约 2 年前

1 comment

pabo大约 2 年前
I was really hoping someone would reproduce the tests and validate the claims from Microsoft&#x27;s &quot;Sparks of AGI&quot; paper [1]. This video just does that (and more).<p>Let me provide one imporant quote from the video (from 18:50):<p>&quot;I personally reproduced every single example from Microsoft, and while all the capabilities of GPT-4 were not necessarily over exaggerated, but the difference between GPT-3.5 and GPT-4 does feel a bit over-inflated [...] one of my quibbles with the Microsoft paper: they give the impression that GPT-4 is an even bigger step for AI than I think is realistically true.&quot;<p>Kudos!<p>[1] <a href="https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2303.12712" rel="nofollow">https:&#x2F;&#x2F;arxiv.org&#x2F;abs&#x2F;2303.12712</a>