Google DeepMind (@GoogleDeepMind) “Here are 2️⃣ ways we’re making sure Gemini is being developed safely and respons” — TopicDigg
Google DeepMind
@GoogleDeepMind
The engine room of @Google. Building AI safely and responsibly to solve the world’s most complex problems. Join us:
加入 January 2016
280 正在关注 1.4M 粉丝
Here are 2️⃣ ways we’re making sure Gemini is being developed safely and responsibly:
⚠️ By using new reinforcement learning methods to improve how models handle sensitive topics.
🛡️ Red teaming to assess security risks posed by indirect prompt injection - a type of cybersecurity attack.
显示更多
