DeepMind
building dialogue agents are all the rage at the moment we have our own powerful systems of course sparrow an information seeking dialogue agent designed to be more helpful correct and designed to answer questions with responses that are useful and evidence based safety is reinforcement learning from human feedback and a set of rules help reduce the risk of unsafe and inappropriate answers frontier fun engaging accurate safe improving alignment of dialogue agents via targeted human | DeepMind
Company
Deck Type
Deck date
April 2023
Slide
24 of 27
Other recent decks by DeepMind
Product
March 2019
Search Thousands of Presentations by World Leading Companies

Stay in the loop

Join our mailing list to stay in the loop with updates and newest feature releases
© 2021-2023 Slidebook.io