web.archive.org/web/20240412043613/http:/news.mit.edu/2024/faster-better-way-preventing-ai-chatbot-toxic-responses-0410

Preview meta tags from the web.archive.org website.

Linked Hostnames

1

Thumbnail

Search Engine Appearance

Google

https://web.archive.org/web/20240412043613/http:/news.mit.edu/2024/faster-better-way-preventing-ai-chatbot-toxic-responses-0410

A faster, better way to prevent an AI chatbot from giving toxic responses

A new technique can more effectively perform a safety check on an AI chatbot. MIT researchers enabled their model to prompt a chatbot to generate toxic responses, which are used to prevent the chatbot from giving hateful or harmful answers when deployed.



Bing

A faster, better way to prevent an AI chatbot from giving toxic responses

https://web.archive.org/web/20240412043613/http:/news.mit.edu/2024/faster-better-way-preventing-ai-chatbot-toxic-responses-0410

A new technique can more effectively perform a safety check on an AI chatbot. MIT researchers enabled their model to prompt a chatbot to generate toxic responses, which are used to prevent the chatbot from giving hateful or harmful answers when deployed.



DuckDuckGo

https://web.archive.org/web/20240412043613/http:/news.mit.edu/2024/faster-better-way-preventing-ai-chatbot-toxic-responses-0410

A faster, better way to prevent an AI chatbot from giving toxic responses

A new technique can more effectively perform a safety check on an AI chatbot. MIT researchers enabled their model to prompt a chatbot to generate toxic responses, which are used to prevent the chatbot from giving hateful or harmful answers when deployed.

  • General Meta Tags

    10
    • title
      A faster, better way to prevent an AI chatbot from giving toxic responses | MIT News | Massachusetts Institute of Technology
    • charset
      utf-8
    • viewport
      width=device-width, initial-scale=1.0
    • description
      A new technique can more effectively perform a safety check on an AI chatbot. MIT researchers enabled their model to prompt a chatbot to generate toxic responses, which are used to prevent the chatbot from giving hateful or harmful answers when deployed.
    • keywords
      Zhang-Wei Hong, James Glass, Pulkit Agrawal, ChatGPT misuses, Toxic AI chatbots, Red-teaming, Chatbot safety, Hateful chatbot, harmful chatbot

  • Open Graph Meta Tags

    7
    • og:title
      A faster, better way to prevent an AI chatbot from giving toxic responses
    • og:image
      https://web.archive.org/web/20240410053813im_/https://news.mit.edu/sites/default/files/images/202404/MIT_Red-Teaming-01-press.jpg
    • og:description
      A new technique can more effectively perform a safety check on an AI chatbot. MIT researchers enabled their model to prompt a chatbot to generate toxic responses, which are used to prevent the chatbot from giving hateful or harmful answers when deployed.
    • og:site_name
      MIT News | Massachusetts Institute of Technology
    • og:type
      article

  • Twitter Meta Tags

    2
    • twitter:site
      @mit
    • twitter:card
      summary_large_image

  • Link Tags

    12
    • apple-touch-icon
      /web/20240410053813im_/https://news.mit.edu/themes/mit/assets/img/favicon/apple-touch-icon.png
    • canonical
      https://web.archive.org/web/20240410053813/https://news.mit.edu/2024/faster-better-way-preventing-ai-chatbot-toxic-responses-0410
    • icon
      /web/20240410053813im_/https://news.mit.edu/themes/mit/assets/img/favicon/favicon.ico
    • icon
      /web/20240410053813im_/https://news.mit.edu/themes/mit/assets/img/favicon/favicon.svg
    • image_src
      https://web.archive.org/web/20240410053813im_/https://news.mit.edu/sites/default/files/images/202404/MIT_Red-Teaming-01-press.jpg

Links

100