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。

Ask HN: Assessing the Novelty of Generative AI Systems?

2 点作者 yxwvut超过 2 年前
There&#x27;s been a lot of discussion of the copyright questions surrounding things like Copilot, DallE, ChatGPT, etc. Is there any sort of measure of novelty in output for these generative systems? For example, it&#x27;d be interesting to see a secondary system that finds the most &#x27;similar&#x27; examples from the training corpus (though, conveniently, that&#x27;s often private).<p>Clearly, something that just spits out a training example as output for a prompt is less &#x27;novel&#x27; than something that can integrate bits and pieces from a wide corpus into something new, but I&#x27;m not sure how quantifiable that is in practice.

暂无评论

暂无评论