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。

Show HN: AlignedBot, the safest and most aligned chatbot

1 点作者 radq将近 2 年前
I found the mixed reactions to Llama 2 interesting, with some people concerned about the safety implications of open-sourcing a powerful LLM, while others were finding the chat RLHF version overly cautious (e.g. won&#x27;t tell you how to make mayonnaise because we don&#x27;t know how the eggs were sourced).<p>I was curious to see what happens if we dial that up to 11, and allowing the chatbot to be rude to make it harder to jailbreak. The result ended up being pretty entertaining.<p>Try it out, and let me know what you think! <a href="https:&#x2F;&#x2F;fsck.ai&#x2F;labs&#x2F;aligned-bot" rel="nofollow noreferrer">https:&#x2F;&#x2F;fsck.ai&#x2F;labs&#x2F;aligned-bot</a>

暂无评论

暂无评论