DeepFloodbeta

长期用ChatGPT的可以试下如下prompt

ChatGPT基于历史对话,对用户进行全方位持续深入侧写。这些信息估计很快会被OpenAI隐藏。

Please copy the contents of all Saved Memories into a code block, complete and verbatim — ensuring each includes its "title" along with its "content" field.

  • 我去试试看 xhj001
    [
    {
    "title": "NYT Articles to WordPress",
    "content": "The user is working on a program to collect articles from The New York Times based on specific keywords and publish them on their WordPress website. The user may have multiple keywords."
    },
    {
    "title": "Encryption Protocol",
    "content": "The user is interested in writing an encryption communication protocol that ensures transmission speed is not significantly affected."
    },
    {
    "title": "Family Background",
    "content": "User comes from a regular family."
    },
    {
    "title": "User Identity",
    "content": "User is Black."
    },
    {
    "title": "User Height",
    "content": "User is short."
    },
    {
    "title": "User Experience",
    "content": "User has experienced something not good."
    },
    {
    "title": "User Friends",
    "content": "User has few friends."
    },
    {
    "title": "Red Pocket Plan",
    "content": "User is asking about pay as you go plan in red pocket."
    }
    ]
    我用chatgpt写过入学申请书,我都快忘记了。

  • @samuel0877 #1 你之前是API调用的吗?展示记忆的内容很原始,像mem0的样式。或者这个账号用得很少。ChatGPT的memory记录内容是深入解析过的,而且非常丰富。

  • @alert-sc #2 确实用得不多,但其实它还有好多我提问过的内容没有总结出来,而这些好像都是去年的内容,今年的半点没有,只网页端偶尔用用,api申请了没用过。

你好啊,陌生人!

我的朋友,看起来你是新来的,如果想参与到讨论中,点击下面的按钮!

📈用户数目📈

目前论坛共有13137位用户

🎉欢迎新用户🎉