Your browser doesn't support javascript.
loading
Learning few-shot imitation as cultural transmission.
Bhoopchand, Avishkar; Brownfield, Bethanie; Collister, Adrian; Dal Lago, Agustin; Edwards, Ashley; Everett, Richard; Fréchette, Alexandre; Oliveira, Yanko Gitahy; Hughes, Edward; Mathewson, Kory W; Mendolicchio, Piermaria; Pawar, Julia; Pȋslar, Miruna; Platonov, Alex; Senter, Evan; Singh, Sukhdeep; Zacherl, Alexander; Zhang, Lei M.
Afiliación
  • Bhoopchand A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Brownfield B; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Collister A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Dal Lago A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Edwards A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Everett R; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Fréchette A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Oliveira YG; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Hughes E; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK. edwardhughes@google.com.
  • Mathewson KW; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Mendolicchio P; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Pawar J; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Pȋslar M; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Platonov A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Senter E; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Singh S; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Zacherl A; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
  • Zhang LM; Google DeepMind, 6-8 Handyside Street, London, N1C 4UZ, UK.
Nat Commun ; 14(1): 7536, 2023 Nov 28.
Article en En | MEDLINE | ID: mdl-38016945
ABSTRACT
Cultural transmission is the domain-general social skill that allows agents to acquire and use information from each other in real-time with high fidelity and recall. It can be thought of as the process that perpetuates fit variants in cultural evolution. In humans, cultural evolution has led to the accumulation and refinement of skills, tools and knowledge across generations. We provide a method for generating cultural transmission in artificially intelligent agents, in the form of few-shot imitation. Our agents succeed at real-time imitation of a human in novel contexts without using any pre-collected human data. We identify a surprisingly simple set of ingredients sufficient for generating cultural transmission and develop an evaluation methodology for rigorously assessing it. This paves the way for cultural evolution to play an algorithmic role in the development of artificial general intelligence.

Texto completo: 1 Colección: 01-internacional Banco de datos: MEDLINE Idioma: En Revista: Nat Commun Asunto de la revista: BIOLOGIA / CIENCIA Año: 2023 Tipo del documento: Article País de afiliación: Reino Unido

Texto completo: 1 Colección: 01-internacional Banco de datos: MEDLINE Idioma: En Revista: Nat Commun Asunto de la revista: BIOLOGIA / CIENCIA Año: 2023 Tipo del documento: Article País de afiliación: Reino Unido