TechHacker Newsabout 3 hours ago
Explorative Modeling boosts generative model efficiency
1 min read
Explorative Modeling trains models on the best of K guesses, improving existing generative models across images, video, and language. Explorative Models reach 6.2x sample efficiency, 4.1x FLOP efficiency, and 47% better parameter efficiency. As end-to-end models, they match diffusion on control tasks with up to 256x less inference compute.
Level
Hype check
Tap to vote and see what everyone thinks.
