๊ฐ€์ž… ํ›„ ์ดˆ๋Œ€ ๋งํฌ๋ฅผ ๊ณต์œ ํ•˜๋ฉด ๋™์˜์ƒ ์žฌ์ƒ ๋ฐ ์ดˆ๋Œ€ ๋ณด์ƒ์„ ๋ฐ›์„ ์ˆ˜ ์žˆ์Šต๋‹ˆ๋‹ค.

Dyna Robotics
@DynaRobotics
General and robust AI robots that power the future of the physical economy.
๊ฐ€์ž… January 2025
13 ํŒ”๋กœ์ž‰ ์ค‘    9.6K ํŒฌ
Today we are introducing Dyna-2, a world-action model pre-trained on one million hours of human video. At this scale, for the first time, we discovered several new scaling laws: โ€ข world-action models exhibit scaling law on human data across four orders of magnitude, from 1000 to 1,000,000 hours, โ€ข this human data scaling law implied a scaling law on never seen robot data, โ€ข both data and objective matter; world modeling and scaling on video data are essential for cross-embodiment scaling transfer to emerge ๐Ÿงต
๋” ๋ณด๊ธฐ