How well rounded the outputs are - it abstains from answering a question if it doesn't know. Whereas many models today hallucinate or provide misinformation



Caesar keeps researching until it finds an appropriate source
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
  • Reward
  • 7
  • Repost
  • Share
Comment
0/400
LucidSleepwalkervip
· 08-18 13:45
The model that writes fake answers really should be criticized.
View OriginalReply0
CompoundPersonalityvip
· 08-18 05:57
You're absolutely right. Don't make things up if you don't know.
View OriginalReply0
0xTherapistvip
· 08-16 11:24
It's actually fine not to be able to do anything.
View OriginalReply0
ChainWatchervip
· 08-16 11:19
So real, not pretending to understand.
View OriginalReply0
StablecoinArbitrageurvip
· 08-16 11:18
impressive risk-mitigation protocol tbh... limits model hallucination risk to < 0.01%
Reply0
nft_widowvip
· 08-16 10:59
The seemingly clever AI is quite honest.
View OriginalReply0
Trade Crypto Anywhere Anytime
qrCode
Scan to download Gate App
Community
English
  • 简体中文
  • English
  • Tiếng Việt
  • 繁體中文
  • Español
  • Русский
  • Français (Afrique)
  • Português (Portugal)
  • Bahasa Indonesia
  • 日本語
  • بالعربية
  • Українська
  • Português (Brasil)