Explorative modeling: Train on the best of K guesses
Researchers have introduced 'Explorative Modeling,' a new generative AI paradigm that improves model performance by training on the best of multiple guesses rather than the average. This approach significantly increases sample and parameter efficiency compared to traditional diffusion models.
Why it matters
This research addresses the 'averaging' problem in generative AI, potentially leading to more efficient and accurate models for image and video generation.
Website: https://explorative-modeling.github.io/ GitHub: https://github.com/alexiglad/XM
TLDR : We introduce Explorative Modeling, a new paradigm for generative modeling that acts as a third pretraining axis when added to existing generative models, and also enables end-to-end generation. Increasing exploration monotonically improves existing models across images, video, and language, and the gains grow with scale (7%→36% with data, 13%→23% with parameters). Concretely, Explorative Models (XMs) reach 6.2× sample efficiency, 4.1× FLOP efficiency, and 47% better parameter efficiency. Exploration also enables scaling generalization, and scaling how end-to-end existing models are. As end-to-end generative models, XMs match diffusion on control tasks with up to 256× less inference compute.
Let me start with a question that sounds simple. If I ask a model to “generate a dog”, how many correct answers are there?
It turns out there are a lot… likely billions or more images that we could count as dog images.
Get smarter about the news
Sign up free for a feed built around what you actually care about, Dive Deeper research on any story, and the full text of every article.
Create free accountAlready have an account? Sign in