RT-2

GPTKB entity

Statements (32)
Predicate Object
gptkbp:instanceOf robotics model
gptkbp:application robotics
robot control
vision-language-action modeling
gptkbp:basedOn transformer architecture
gptkbp:canGeneralizeTo unseen objects
unseen tasks
gptkbp:demonstrates gptkb:Google_DeepMind
robotic arms
gptkbp:developedBy gptkb:Google_DeepMind
gptkbp:fullName Robotic Transformer 2
https://www.w3.org/2000/01/rdf-schema#label RT-2
gptkbp:input gptkb:text
images
gptkbp:language English
gptkbp:notableFor robotic manipulation
zero-shot generalization
end-to-end vision-language-action learning
gptkbp:notablePublication RT-2: Vision-Language-Action Models Transfer Web Knowledge to Robotic Control
https://arxiv.org/abs/2307.15818
gptkbp:openSource no
gptkbp:output robot actions
gptkbp:predecessor gptkb:RT-1
gptkbp:relatedTo transformer models
multimodal AI
robotics foundation models
gptkbp:releaseDate 2023
gptkbp:uses vision-language models
large language model techniques
gptkbp:bfsParent gptkb:Soviet_missile_program
gptkb:A_Generalist_Agent
gptkbp:bfsLayer 7