DeepMind’s Proposes New Paradigm for Interfacing Language Model with Robots Through Rewards

In a new paper Language to Rewards for Robotic Skill Synthesis, a Google DeepMind research team proposes a new paradigm to leverage reward functions to interface language and low-level robot actions, which enables non-technical users to steer novel and intricate robot actions without large amount of data or expert knowledge to engineer low-level primitives.

ai artificial intelligence data deepmind deepmind research deep-neural-networks functions google google deepmind language language model large language model low machine learning machine learning & data science ml new paradigm novel paper paradigm research research team robot robotics robots synthesis team technical technology through transformers